Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
warwick.ac.uk/lib-publications
A Thesis Submitted for the Degree of PhD at the University of Warwick
Permanent WRAP URL:
http://wrap.warwick.ac.uk/89468
Copyright and reuse:
This thesis is made available online and is protected by original copyright.
Please scroll down to view the document itself.
Please refer to the repository record for this item for information to help you to cite it.
Our policy information is available from the repository home page.
For more information, please contact the WRAP Team at: [email protected]
Antibiotic Biosynthesis and its Transcriptional
Regulation in Streptomyces Bacteria
Shanshan Zhou
Supervisor: Prof. Gregory L. Challis
A thesis submitted in partial fulfillment of the requirements for the degree of Doctor of
Philosophy in Chemistry
Department of Chemistry, University of Warwick
September 2016
i
Table of Contents
List of Figures ............................................................................................................................. v
List of Schemes ........................................................................................................................ xiii
List of Tables ........................................................................................................................... xvi
Acknowledgements ................................................................................................................ xvii
Declaration ............................................................................................................................ xviii
Abstract .................................................................................................................................... xix
Abbreviations ........................................................................................................................... xx
1. Introduction ........................................................................................................................ 1
1.1 Streptomyces ........................................................................................................................ 2
1.1.1 Streptomyces coelicolor A3(2) ................................................................................. 5
1.2 Streptomyces signalling molecules ...................................................................................... 7
1.2.1 Structural types of Streptomyces signalling molecules ............................................ 7
1.2.2 Biosynthesis of γ-butyrolactones ........................................................................... 10
1.2.3 Biosynthesis of AHFCAs ....................................................................................... 14
1.3 Transcriptional regulation of specialised metabolites in Streptomyces ............................. 17
1.3.1 TetR transcriptional repressors .............................................................................. 17
1.3.2 ArpA-like transcriptional repressors ...................................................................... 18
1.3.3 Regulation of methylenomycin biosynthesis in S. coelicolor ................................ 20
1.4 Siderophores ...................................................................................................................... 24
1.4.1 Bioactivity of siderophores .................................................................................... 26
1.4.2 Biosynthesis of siderophores ................................................................................. 26
1.5 2-Hydroxyphenothiazolines ............................................................................................... 27
1.5.1 Pyochelin and enantiopyochelin ............................................................................ 29
1.5.2 Biosynthesis of pyochelin and enantio-pyochelin .................................................. 31
1.5.3 Pulicatins, thiazostatin and watasemycin ............................................................... 35
1.6 S-Adenosylmethionine (SAM)-dependent methylation ..................................................... 36
1.6.1 Class B radical SAM methylases ........................................................................... 38
1.6.1.1 Stereochemical course of C-methylation by class B radical SAM methylases .. 40
1.7 Aims of project .................................................................................................................. 42
2. Investigation of the role of MmfL in AHFCA biosynthesis .......................................... 43
2.1 Introduction to previous studies ......................................................................................... 44
2.2 Synthesis of the N-acetylcysteamine (NAC) β-ketothioester substrate analogue .............. 46
2.3 Synthesis of the butenolide authentic standard .................................................................. 47
2.4 Overproduction, purification and characterisation of recombinant MmfL ........................ 51
ii
2.5 Characterisation of dephosphorylated product of MmfL-catalysed reaction ..................... 52
2.6 Summary ............................................................................................................................ 56
3. Interactions between DNA, ligand and ArpA-like repressors ..................................... 57
3.1 Introduction ........................................................................................................................ 58
3.2 Interactions of MmfR with three MARE sequences .......................................................... 60
3.2.1. Confirmation of MmfR binding to MARE sequences ........................................... 60
3.2.2. Comparison of the binding affinity of MmfR to three MAREs ............................. 64
3.3 Interactions of MmfR with natural AHFCAs .................................................................... 67
3.3.1 Synthesis of AHFCA1 and AHFCA3 .................................................................... 67
3.3.2 Comparison of the binding affinity of the different intergenic regions to MmfR
with natural AHFCAs 1-5 ...................................................................................................... 70
3.4 Key residues in the ligand binding of AHFCAs by MmfR ................................................ 72
3.4.1 Analysis of the crystal structure ............................................................................. 72
3.4.2 Preparation of MmfR mutants ................................................................................ 73
3.4.3 Attempts to mutate Leu 110 ................................................................................... 75
3.4.4 Effect of mutations on ligand binding .................................................................... 76
3.4.4.1. Tyr 85 ............................................................................................................. 76
3.4.4.2. Gln 130........................................................................................................... 77
3.4.4.3. Tyr 144 ........................................................................................................... 78
3.5 Ligand binding specificity of MmfR ................................................................................. 78
3.5.1 Synthesis of SCB1 ................................................................................................. 79
3.5.2 Interaction of MmfR with SCB1 ............................................................................ 83
3.6 Investigation of a MmfR homologue, SgnR ...................................................................... 84
3.6.1 Cloning of sgnR ..................................................................................................... 85
3.6.2 Overproduction and purification of recombinant His6-SgnR protein .................... 86
3.6.3 Synthesis of AHFCA6 and AHFCA7 .................................................................... 87
3.6.4 EMSAs with SgnR ................................................................................................. 88
3.7 Key notes for studying in vitro DNA-protein-ligand interaction using EMSAs ................ 90
3.8 Summary ............................................................................................................................ 92
4. Watasemycin biosynthesis in Streptomyces venezuelae ATCC10712 ........................... 94
4.1 Introduction to previous work ............................................................................................ 95
4.1.1 Upregulation of ‘pyochelin-like’ gene cluster ....................................................... 95
4.1.2 Isolation and characterisation of watasemycin and thiazostatin from S. venezuelae .
............................................................................................................................... 97
4.2 Identification and characterisation of isopyochelin ......................................................... 100
4.2.1 Synthesis of pyochelin ......................................................................................... 100
4.2.2 LC-MS analysis of synthetic pyochelin ............................................................... 104
4.2.3 Synthesis and characterisation of isopyochelin .................................................... 105
iii
4.3 Roles of Sven0508 and Sven0516 in the biosynthetic pathway ...................................... 108
4.3.1 Sven0516 is required for 2-hydroxyphenylthiazoline biosynthesis ..................... 109
4.3.2 Sven0508 is not involved in biosynthesis of 2-hydroxyphenylthiazolines .......... 110
4.4 Reassignment of the absolute stereochemistry of watasemycin and thiazostatin ............ 111
4.5 Role of Sven0515 in watasemycin biosynthesis .............................................................. 113
4.5.1 Sven0515 methylates C-5' of thiazostatin to produce watasemycin .................... 113
4.5.2 Synthesis of stereospecifically deuterium-labelled cysteine ................................ 115
4.5.3 Incorporation of stereospecifically deuterium-labelled L-cystine into watasemycin .
............................................................................................................................. 123
4.6 Proposed pathway for watasemycin biosynthesis ............................................................ 129
4.7 Summary .......................................................................................................................... 132
5. Conclusions and future work ........................................................................................ 134
5.1 MmfL catalyses the formation of a butenolide intermediate in AHFCA biosynthesis .... 135
5.2 Interactions between DNA, ligand and ArpA-like repressors ......................................... 136
5.2.1 Three operators for MmfR binding ...................................................................... 136
5.2.2 Molecular basis for AHFCA recognition by MmfR ............................................ 136
5.2.3 MmfR is not responsive to γ-butyrolactones ....................................................... 137
5.2.4 SgnR in the regulation of gaburedin biosynthesis................................................ 137
5.3 Watasemycin biosynthesis in S. venezuelae ATCC10712 ............................................... 138
5.3.1 Isopyochelin is a shunt metabolite in the watasemycin biosynthetic pathway .... 138
5.3.2 Sven0516 is the only thiazoline reductase required for watasemycin biosynthesis ...
............................................................................................................................. 139
5.3.3 Sven0515 catalyses C-methylation of thiazostatin to produce watasemycin ....... 139
5.3.4 Stereochemistry of thiazostatin and watasemycin ............................................... 141
6. Experimental .................................................................................................................. 143
6.1 Synthetic chemistry .......................................................................................................... 144
6.1.1 General information ............................................................................................. 144
6.1.2 Synthesis of monosilyl DHA ............................................................................... 147
6.1.3 Synthetic MmfL substrate and enzymatic product ............................................... 148
6.1.3.1 Synthesis of NAC β-ketothioester.................................................................... 148
6.1.3.2 Synthetic enzymatic product ............................................................................ 150
6.1.4 Synthesis of SCB6 ............................................................................................... 151
6.1.5 Synthesis of 6-methylheptanoic acid ................................................................... 154
6.1.6 Synthesis of SCB1 ............................................................................................... 156
6.1.7 Synthesis of AHFCAs .......................................................................................... 159
6.1.7.1 AHFCA1 .......................................................................................................... 159
6.1.7.2 AHFCA3 .......................................................................................................... 160
6.1.7.3 AHFCA6 .......................................................................................................... 162
iv
6.1.7.4 AHFCA7 .......................................................................................................... 164
6.1.8 Synthesis of pyochelin and isopyochelin ............................................................. 166
6.1.9 Synthesis of stereospecifically deuterium-labelled L-cystine .............................. 169
6.2 Analytical chemistry ........................................................................................................ 178
6.2.1 Extraction protocol ............................................................................................... 178
6.2.2 Liquid chromatography-mass spectrometry (LC-MS) ......................................... 178
6.2.3 Chiral LC-MS analysis......................................................................................... 179
6.2.4 Circular dichroism (CD) spectroscopy ................................................................ 179
6.3 Biology ............................................................................................................................. 180
6.3.1 Bacterial strains .................................................................................................... 180
6.3.2 Culture media ....................................................................................................... 180
6.3.3 Preparation of bacterial spores ............................................................................. 180
6.3.4 Production of 2-hydroxyphenylthiazoline metabolites ........................................ 181
6.3.5 Feeding the extract from ∆sven0515 mutant to ∆sven0516 mutant ..................... 181
6.3.6 Feeding experiment with deuterium-labelled precursors ..................................... 181
6.3.7 Chemical transformations .................................................................................... 182
6.3.8 Plasmid DNA preparation .................................................................................... 182
6.3.9 Restriction digests ................................................................................................ 183
6.3.10 Polymerase Chain Reaction (PCR) ...................................................................... 183
6.3.11 Agarose gel electrophoresis ................................................................................. 184
6.3.12 Purification of DNA from agarose gels ............................................................... 185
6.3.13 Construction of plasmid pET151-sgnR ................................................................ 185
6.3.14 Site-directed mutagenesis of MmfR .................................................................... 186
6.3.15 Expression and overproduction of recombinant protein ...................................... 187
6.3.16 Purification of recombinant protein ..................................................................... 187
6.3.17 Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) ........ 188
6.3.18 Electrophoretic mobility shift assay (EMSA) ...................................................... 189
6.3.19 Enzymatic assay of MmfL ................................................................................... 190
References ............................................................................................................................... 192
v
List of Figures
Figure 1.1. The Streptomyces life cycle.
Figure 1.2. Examples of bioactive metabolites that are produced by Streptomyces species.
The activity and producing organism for each metabolite is shown below its name.
Figure 1.3. Representative structures of specialised metabolites produced by S.
coelicolor.
Figure 1.4. Three structural types of GBLs.
Figure 1.5. Structures of Streptomyces GBL-type signalling molecules. The absolute
stereochemistry of factor I and Gräfe factors are not clear.
Figure 1.6. Structures of Streptomyces AHFCA- and butenolide-type signalling
molecules. SRB1 and SRB2 exist as mixtures of interconverting epimers at C-4.
Figure 1.7. Incorporation results using 13C-labelled glycerol, acetate and isovalerate for
investigation of VB-A 33 biosynthesis by Sakuda and co-workers.
Figure 1.8. Alignment of the amino acid sequences of AfsA from S. griseus and the S.
coelicolor AfsA homologues ScbA and MmfL. Identical and conservatively substituted
amino acids are highlighted with asterisk and dots, respectively.
Figure 1.9. An example of the feeding experiments using deuterium-labelled precursors
to yield the AHFCAs.
Figure 1.10. Mode of action of TetR. TetR repressor binds to the tetA promoter and
represses tetA expression. The binding of tetracycline to TetR releases it from the
promoter, thus allowing transcription of tetA. The TetA protein exports tetracycline out
of the cell.
Figure 1.11. The A-Factor/ArpA regulatory cascade in S. griseus leading to transcription
of adpA and subsequent morphological development and production of specialised
metabolites, such as streptomycin and grixazone.
Figure 1.12. Organisation of the methylenomycin biosynthetic gene cluster in S.
coelicolor A3(2).
Figure 1.13. X-ray crystal structure of MmfR with AHFCA2 bound.
Figure 1.14. Proposed signalling cascade with MmfR involved in methylenomycin
biosynthesis. MARE sequences for MmfR binding are located in the intergenic regions
of mmfL-mmfR, mmyY-mmyB and the region upstream of mmyR. MmyR is also proposed
to bind the same MAREs.
vi
Figure 1.15. Location of AREs, which are shown in purple, in the S. venezuelae
gaburedin biosynthetic gene cluster.
Figure 1.16. Representative examples of microbial siderophores.
Figure 1.17. Representative members of the 2-hydroxyphenylthiazoline family
compounds found in bacteria.
Figure 1.18. Proposed mechanism for epimerisation at C-2 of 2,4-disubstituted
thiazolidine compounds.
Figure 1.19. Organisation of the pyochelin biosynthetic gene clusters in P. protegens
CHA0, P. aeruginosa PAO1 and S. scabies 87-22. Pyochelin biosynthetic gene
homologues are colour-coded based on conserved function. Genes presumably involved
in pyochelin export are grey, while non-homologous genes are white. The pchE gene is
smaller in P. protegens CHA0 than in P. aeruginosa PAO1 due to the absence of an MT-
like coding region (coloured in black in pchE in P. aeruginosa PAO1).
Figure 1.20. Proposed pathway for the biosynthesis of pyochelin in P. aeruginosa.
Figure 1.21. Proposed pathway for the biosynthesis of enantio-pyochelin in P. protegens.
Figure 1.22. Reported relative stereochemistry of watasemycin and thiazostatin.
Figure 1.23. Three typical SAM-dependent reactions with methyl, 3-amino-3-
carboxypropyl or adenosyl group transfer.
Figure 1.24. Generation of 5′-dAdo radical intermediate from SAM as an initiation step
for methylation reactions catalysed by RS methylases.
Figure 1.25. Specialised metabolites for which class B RS methylases are known or
predicted to be involved in their biosynthesis, with corresponding functional groups
coloured in red.
Figure 1.26. Proposed catalytic mechanism of class B RS methylases (left) and the
structure of MeCbl (right).
Figure 1.27. Proposed organisation of the active site in class B RS methylases.
Figure 2.1. EICs at m/z = 235.0941 (A and B), corresponding to [M+Na]+ for butenolide
107, (A) from LC-MS analysis of the deprotection by hydrogen fluoride-pyridine, (B)
from a sample of the synthetic TBDMS-protected butenolide 106, and m/z = 349.1806
(C), corresponding to [M+Na]+ for TBDMS-protected butenolide 106.
Figure 2.2. Positive ion mode ESI-MS/MS spectrum of the synthetic butenolide 107
from deprotection by hydrogen fluoride-pyridine complex with predicted fragmentation
annotated.
vii
Figure 2.3. SDS-PAGE analysis of His6-MmfL (~41 kDa) after purification. (MWM:
molecular weight marker)
Figure 2.4. Measured (top) and deconvoluted (bottom) mass spectra of His6-MmfL
(calculated mass = 41442.0 Da).
Figure 2.5. EICs for m/z = 235.0941 (corresponding to [M+Na]+ for butenolide 107)
from LC-MS analyses of the dephosphorylated product of the MmfL-catalysed reaction
of DHAP with NAC β-ketothioester 101.
Figure 2.6. EICs for m/z = 235.0941 from LC-MS analyses of the dephosphorylated
product of the MmfL-catalysed reaction of DHAP with NAC β-ketothioester 101 and
synthetic AHFCA5 44, as well as a co-injection of them.
Figure 2.7. (A) EICs for m/z = 235.0941 from LC-MS analyses of the dephosphorylated
product of the MmfL-catalysed reaction of DHAP with NAC β-ketothioester 101 and
synthetic butenolide 107, as well as a co-injection of them in equal volumes.
(B) Comparison of the MS/MS fragmentation spectra of the peaks with m/z = 213
(corresponding to [M+H]+) obtained from synthetic butenolide 107 and the enzymatic
reaction product. Precursor ions are indicated with diamonds.
Figure 3.1. Schematic representation of EMSAs in this study.
Figure 3.2. An example of the EMSAs with DNA fragment of the entire mmfR-mmfL
intergenic region (194 bp), MmfR and AHFCA1 by P. Harrison (left), and the S.
coelicolor methylenomycin biosynthetic gene cluster with location of three proposed
MAREs highlighted in magenta (right).
Figure 3.3. SDS-PAGE analysis of His6-MmfR (~27.8 kDa) after purification. (MWM:
molecular weight marker)
Figure 3.4. Measured (top) and deconvoluted (bottom) mass spectra of His6-MmfR
(calculated mass = 27835.5 Da).
Figure 3.5. Hairpin DNA containing three MAREs (18 bp, coloured in red) for EMSAs.
Figure 3.6. Interaction of MmfR with a self-annealing DNA fragment forming a hairpin
sequence containing MARE1 located in the mmfL-mmfR intergenic region in response to
increasing amounts of AHFCA1. Lane 1: DNA fragment only. Lanes 2 to 7: AHFCA1
at 0, 0.8, 8, 20, 40 and 100 nmol respectively. Amounts of DNA and proteins were kept
constant (0.8 pmol and 4.0 pmol respectively).
Figure 3.7. Interaction of MmfR with self-annealing DNA fragments forming hairpin
sequences containing MARE2 located in the mmyY-mmyB intergenic region (left) or
MARE3 located upstream of mmyR (right) in response to increasing amounts of
viii
AHFCA1. Lane 1: DNA fragment only. Lanes 2 to 7: AHFCA1 at 0, 0.8, 8, 20, 40 and
100 nmol, respectively. Amounts of DNA and proteins were kept constant (0.8 pmol and
4.0 pmol respectively).
Figure 3.8. Agarose gel electrophoresis analyses of the PCR products for different length
of DNA fragments containing MAREs. (LR: low range DNA ladder)
Figure 3.9. Competitive binding of MmfR with two DNA fragments corresponding to
the mmfR-mmfL and mmyY-mmyB intergenic regions. Lanes 1 to 3: DNA fragments only.
Lanes 4 to 9: MmfR at 0.1, 0.2, 0.3, 0.4, 0.5 and 0.6 pmol, respectively. The amount of
DNA was kept constant (0.05 pmol).
Figure 3.10. Competitive binding of MmfR with two DNA fragments corresponding to
the mmfR-mmfL intergenic region and upstream of mmyR. Lanes 1 to 3: DNA fragments
only. Lanes 4 to 9: MmfR at 0.1, 0.2, 0.3, 0.4, 0.5 and 0.6 pmol, respectively. The amount
of DNA was kept constant (0.05 pmol).
Figure 3.11. Competitive binding of MmfR with two DNA fragments corresponding to
the mmyY-mmyB intergenic region and upstream of mmyR. Lanes 1 to 3: DNA fragments
only. Lanes 4 to 9: MmfR at 0.1, 0.2, 0.3, 0.4, 0.5 and 0.6 pmol, respectively. The amount
of DNA was kept constant (0.05 pmol).
Figure 3.12. Interaction of MmfR with the 194-bp DNA fragment of the entire mmfR-
mmfL intergenic region in response to increasing amounts of natural AHFCAs 1-5. Lane
1: DNA fragment only. Lanes 2 to 9: AHFCAs at 0, 0.8, 4, 8, 14, 20, 40 and 100 nmol,
respectively. Amounts of DNA and proteins were kept constant (0.1 pmol and 1.8 pmol
respectively).
Figure 3.13. Interaction of MmfR with the 230-bp DNA fragment of the entire mmyY-
mmyB intergenic region in response to increasing amounts of natural AHFCAs 1-5. Lane
1: DNA fragment only. Lanes 2 to 9: AHFCAs at 0, 0.8, 4, 8, 14, 20, 40 and 100 nmol,
respectively. Amounts of DNA and proteins were kept constant (0.1 pmol and 1.8 pmol
respectively).
Figure 3.14. (A) Proposed key residues of MmfR interacting with AHFCA2 from solved
crystal structure. (B) Key hydrogen bonding interactions in the hydrophilic part of
AHFCA2.
Figure 3.15. SDS-PAGE analysis of His6-MmfR mutants, Y85F, Y85A, Q130E, Y144F
and Y144A, after purification. (MWM: molecular weight marker)
Figure 3.16. Measured (top) and deconvoluted (bottom) mass spectra of His6-MmfR
mutants, Y85F (calculated mass = 27819.5 Da), Y85A (calculated mass = 27743.5 Da),
ix
Y144F (calculated mass = 27819.5 Da), Y144A (calculated mass = 27743.5 Da) and
Q130E (calculated mass = 27836.5 Da).
Figure 3.17. SDS-PAGE analysis in the attempt to purify the MmfR mutant L110V.
Figure 3.18. Comparison of the effect of AHFCA1 on protein:DNA (mmfL-mmfR
fragment, 194 bp) dissociation when WT MmfR, MmfR Y85F and Y85A were used.
Lane 1: DNA fragment only. Lanes 2 to 9: AHFCA1 at 0, 0.8, 4, 8, 14, 20, 40 and 100
nmol, respectively. Amount of DNA and proteins were kept constant (0.1 pmol and 1.8
pmol respectively).
Figure 3.19. Comparison of the effect of AHFCA1 on protein:DNA (mmfL-mmfR
fragment, 194 bp) dissociation when WT MmfR and MmfR Q130E mutant were used.
Lane 1: DNA fragment only. Lanes 2 to 9: AHFCA1 at 0, 0.8, 4, 8, 14, 20, 40 and 100
nmol, respectively. Amount of DNA and proteins were kept constant (0.1 pmol and 1.8
pmol respectively).
Figure 3.20. Comparison of the effect of AHFCA1 on protein:DNA (mmfL-mmfR
fragment, 194 bp) dissociation when WT MmfR, MmfR Y144A and Y144F were used.
Lane 1: DNA fragment only. Lanes 2 to 9: AHFCA1 at 0, 0.8, 4, 8, 14, 20, 40 and 100
nmol, respectively. Amount of DNA and proteins were kept constant (0.1 pmol and 1.8
pmol respectively).
Figure 3.21. Comparison of the 400 MHz 1H NMR spectra of racemic SCB1 and their
diastereoisomers.
Figure 3.22. Interaction of MmfR with the 194-bp DNA fragment of the entire mmfR-
mmfL intergenic region in response to increasing amounts of AHFCA1 and racemic
SCB1. Lane 1: DNA fragment only. Lanes 2 to 7: AHFCA1 or racemic SCB1 at 0, 0.8,
8, 20, 40 and 400 nmol, respectively. Amounts of DNA and proteins were kept constant
(0.1 pmol and 1.6 pmol respectively).
Figure 3.23. Agarose gel electrophoresis analysis of the sgnR PCR products (604 bp).
(LR: low range DNA ladder)
Figure 3.24. Agarose gel electrophoresis analysis of restriction digests (left) using PstI
to confirm the constructed pET151-sgnR plasmid, and PCRs (right) using the constructed
pET151-sgnR plasmid as template (604 bp). (LR: low range DNA ladder)
Figure 3.25. SDS-PAGE analysis of total (PTOT) and soluble (PSol) protein fractions
resulting from overproduction of His6-SgnR in E. coli BL21star (DE3) (left), and His6-
SgnR after purification (right). (MWM: molecular weight marker)
x
Figure 3.26. Measured (top) and deconvoluted (bottom) mass spectra of His6-SgnR
(calculated mass = 26159.5 Da).
Figure 3.27. Interaction of SgnR with the 194-bp DNA fragment of the entire mmfR-
mmfL intergenic region. Lane 1: DNA fragment only. Lanes 2 and 3: SgnR at 0.9 and 1.8
pmol, respectively. Amount of DNA was kept constant (0.1 pmol).
Figure 3.28. Agarose gel electrophoresis analysis of the PCR products of the entire sgnR-
sgnL intergenic region (121 bp). (LR: low range DNA ladder)
Figure 3.29. Interaction of SgnR with the 121-bp DNA fragment of the entire sgnR-sgnL
intergenic region in response to increasing amounts of AHFCA6. Lane 1: DNA fragment
only. Lanes 2 to 7: AHFCA6 at 0, 0.8, 8, 20, 40 and 400 nmol, respectively. Amounts of
DNA and proteins were kept constant (0.1 pmol and 1.8 pmol, respectively).
Figure 3.30. Interaction of SgnR with the 28-bp annealed DNA fragments containing
ARE2 located upstream of gbnA (left) or ARE3 located in the gbnR-sgnH intergenic
region (right) in response to increasing amounts of AHFCA6. Lane 1: DNA fragment
only. Lanes 2 to 7: AHFCA6 at 0, 0.8, 8, 20, 40 and 400 nmol, respectively. Amounts of
DNA and proteins were kept constant (0.8 pmol and 4.0 pmol respectively).
Figure 3.31. An example of protein-DNA-ligand reaction with precipitation of the
protein in presence of the compound at a final concentration of 200 mM.
Figure 4.1. Microarray expression profiles of the cluster of genes, sven0503-sven0517,
in S. venezuelae. The y axis shows normalised transcript abundance.
Figure 4.2. Organisation of the pyochelin and watasemycin biosynthetic gene clusters in
S. scabies 87-22 and S. venezuelae ATCC10712, respectively. Genes that encode
functionally analogous proteins are annotated in the same colour.
Figure 4.3. Proposed structures of 2-hydroxyphenylthiazolines produced from the
sven0503-sven0517 gene cluster in S. venezuelae ATCC10712 based on high resolution
LC-MS analyses.
Figure 4.4. Extracted ion chromatograms (EICs) at m/z = 353.0988, 339.0832, 210.0583,
224.0740 and 325.0675, corresponding to [M+H]+ for thiazostatin, watasemycin,
aerugine, pulicatins A/B and pyochelin, respectively, from LC-HRMS analyses of the
ethanol extract of S. coelicolor M1152/SV-2_E03::SspI culture supernatant (blue traces),
in comparison to wild type S. coelicolor M1152 (red traces). Further work in this study
would give more information about the peaks corresponding to [M+H]+ for pyochelin
(see Section 4.2).
Figure 4.5. The structure of synthetic pyochelin and neopyochelin.
xi
Figure 4.6. EICs at m/z = 325.0675, corresponding to [M+H]+ for the metabolite (top
trace) from LC-HRMS analyses of the ethanol extract of S. coelicolor
M1152/SV-2_E03::SspI culture supernatant, and synthetic mixture of pyochelins and
neopyochelins (bottom trace), including two stereoisomers identical to natural pyochelin
from S. scabies 87-22 and P. aeruginosa.
Figure 4.7. EICs for m/z = 325.0675, corresponding to [M+H]+ for isopyochelin, from
LC-HRMS analyses of the ethanol extract of S. coelicolor M1152/SV-2_E03::SspI
culture supernatant (top trace), the authentic synthetic standard of the isopyochelin
stereoisomers from 2-methyl-L-cysteine (middle trace) and the extract to which an
approximately equimolar quantity of the synthetic standard has been added (bottom
trace).
Figure 4.8. CD spectra of commercially available 2-methyl-L-cysteine and 2-methyl-D-
cysteine used in the synthesis of isopyochelin.
Figure 4.9. EICs at m/z = 325.0 (corresponding to [M+H]+ for isopyochelin) from
LC-HRMS analyses of the synthetic stereoisomers of isopyochelin and the natural
metabolites from the ethanol extract of S. coelicolor M1152/SV-2_E03::SspI culture
supernatant on a homochiral stationary phase. Comparison with 4"R-isopyochelins 135
(left) and 4"S-isopyochelins 136 (right) is shown separately.
Figure 4.10. EICs at m/z = 353.0988, 339.0832, 210.0583, 224.0740 and 325.0675,
corresponding to [M+H]+ for watasemycin, thiazostatin, aerugine, pulicatins A/B and
isopyochelin, respectively, from LC-HRMS analyses of the ethanol extract of the
∆sven0516 mutant culture supernatant (purple traces), in comparison to the S. coelicolor
M1152/SV-2_E03::SspI strain (blue traces).
Figure 4.11. EICs at m/z = 353.0988, 339.0832, 210.0583, 224.0740 and 325.0675,
corresponding to [M+H]+ for watasemycin, thiazostatin, aerugine, pulicatins A/B and
isopyochelin, respectively, from LC-HRMS analyses of the ethanol extract of the
∆sven0508 mutant culture supernatant.
Figure 4.12. Two reported possible absolute configurations of watasemycin and
thiazostatin. The absolute stereochemistry with 4"S configuration is circled in red.
Figure 4.13. Reassigned stereochemistry of thiazostatin, watasemycin and isopyochelin.
Figure 4.14. EICs at m/z = 353.0988, 339.0832, 210.0583, 224.0740 and 325.0675,
corresponding to [M+H]+ for watasemycin, thiazostatin, aerugine, pulicatins A/B and
isopyochelin, respectively, from LC-MS analyses of the ethanol extract of the ∆sven0515
xii
culture supernatant (green traces), in comparison to the S. coelicolor
M1152/SV-2_E03::SspI strain (blue traces).
Figure 4.15. EICs from LC-MS analyses of the ethanol extract from the ∆sven0516
mutant fed with the extract from the ∆sven0515 mutant.
Figure 4.16. Feeding of stereospecifically deuterium-labelled substrates to determine the
stereochemical course of the C-methylation reaction in fosfomycin biosynthesis.
Figure 4.17. Feeding of stereospecifically isotope-labelled amino acids to determine the
stereochemical course of the methylations in bottromycin biosynthesis, reported by
Kellenberger and Arigoni.
Figure 4.18. A comparison between 4.20-5.70 ppm of the 1H NMR spectra in CDCl3 at
600 MHz of (3R)-O-acetyl-N-(tert-butoxycarbonyl)-L-serine-3-2H 146 at varying
temperatures.
Figure 4.19. The 1H (at 300 MHz) and 13C NMR (at 500 MHz) spectra of the synthesised
(2R,2'R,3S,3'S)-[3,3'-2H2]cystine 151 in 2.5% NaOD in D2O.
Figure 4.20. A comparison of EICs for thiazostatin (red traces) and watasemycin (black
traces) from the LC-MS analyses of the ethanol extracts of S. coelicolor
M1152/SV-2_E03::SspI culture supernatant grown in SMM and YD medium.
Figure 4.21. Possible deuterium-labelled isotopomers of thiazostatin and watasemycin
produced after feeding with (2R, 3S)-[3-2H1] cysteine 139.
Figure 4.22. A comparison of the HRMS for thiazostatin and watasemycin from LC-MS
analyses of the ethanol extracts of S. coelicolor M1152/SV-2_E03::SspI fed with
(2R,2'R,3S,3'S)-[3,3'-2H2]cystine.
Figure 4.23. An example of deciphering the HRMS of the metabolic products from
feeding experiments with deuterium-labelled precursors, including the isotopic
distribution of natural thiazostatin (top) and singly deuterium-labelled thiazostatin
(middle) and the actual mass spectrum of thiazostatin detected (bottom).
Figure 4.24. Proposed stereochemical course during Sven0515-catalysed methylation at
C-5' of thiazostatin based on the feeding experiment.
Figure 4.25. Proposed pathway for the biosynthesis of thiazostatin 82, watasemycin 83,
isopyochelin 137, aerugine 75 and pulicatins A/B 77/78 in S. venezuelae ATCC 10712.
Figure 5.1. Sequence alignment of SgnR and MmfR. The key amino acid residues
identified from MmfR are highlighted in yellow.
xiii
List of Schemes
Scheme 1.1. Biosynthetic route to A-factor as proposed by Kato and co-workers.
Scheme 1.2. Proposed pathway for the biosynthesis of AHFCAs, taking AHFCA1 as an
example.
Scheme 1.3. Summary of the in vitro enzymatic studies on AHFCA biosynthesis by N.
Malet.
Scheme 2.1. Proposed biosynthetic pathway for AHFCAs, involving different
β-ketoacyl-ACP thioesters.
Scheme 2.2. Enzyme assays performed by Kato and co-workers for investigation of
AfsA function.
Scheme 2.3. Enzymatic reactions performed by N. Malet with synthetic 5-methyl-3-
oxohexanoyl-NAC for investigation of the function of MmfL.
Scheme 2.4. Synthesis of NAC β-ketothioesters 61 and 101 as substrate analogues of
MmfL.
Scheme 2.5. General mechanism for the coupling of the carboxylic acid with Meldrum’s
acid using EDC and DMAP.
Scheme 2.6. Attempted synthesis of the dephosphorylated putative product 60 of the
MmfL-catalysed condensation of DHAP and NAC β-ketothioester 61 by N. Malet.
Scheme 2.7. Synthesis of the TBDMS-protected butenolide 106.
Scheme 2.8. Attempts to deprotect the TBDMS-protected butenolide 106.
Scheme 2.9. MmfL-catalysed condensation of DHAP and NAC β-ketothioester 101,
followed by dephosphorylation to form butenolide 107.
Scheme 2.10. Attempted synthesis of the TBDMS-protected butenolide 106 from the
dephosphorylated product of the MmfL-catalysed condensation of DHAP and NAC
β-ketothioester 101.
Scheme 3.1. Synthesis of AHFCAs reported by Sello and co-workers.
Scheme 3.2. General concept of the Garcia Gonzalez reaction.
Scheme 3.3. Proposed mechanism for Knoevenagel condensation and subsequent furan
formation catalysed by Sc(OTf)3.
Scheme 3.4. Synthesis of the β-ketoesters 110 and 112 for AHFAC1 and AHFCA3,
respectively.
Scheme 3.5. Synthesis of AHFCA1 and AHFCA3.
xiv
Scheme 3.6. Key steps in most reported synthesis of A-factor
Scheme 3.7. Synthesis of SCB1 reported by Takano and co-workers.
Scheme 3.8. Synthesis of A-factor reported by Sello and co-workers.
Scheme 3.9. Synthesis of SCB6 for optimisation purpose.
Scheme 3.10. Synthesis of SCB1 and its stereoisomers in this study.
Scheme 3.11. Synthesis of AHFCA6 and AHFCA7.
Scheme 4.1. First total synthesis of pyochelin reported by Ankenbauer et al.
Scheme 4.2. Improved synthesis of the aldehyde 132 by Rinehart et al.
Scheme 4.3. Improved synthesis of aldehyde 132 by Zamri and co-workers.
Scheme 4.4. Synthetic route to pyochelin used in this study.
Scheme 4.5. The proposed mechanism for CDMT-mediated coupling of acid 131 and
N,O-dimethylhydroxylamine.
Scheme 4.6. Synthesis of a mixture of isopyochelin stereoisomers 135 with R
configuration at C-4".
Scheme 4.7. Synthesis of a mixture of isopyochelin stereoisomers 136 with S
configuration at C-4".
Scheme 4.8. Proposed formation of aerugine 75 and pulicatins A/B 77/78, from
hydrolysis and reduction of isopyochelin/thiazostatin and watasemycin, respectively.
Scheme 4.9. Possible stereochemical courses for Sven0515-catalysed methylation,
which could occur via primarily route A or route B, or via a mixture of routes A and B.
Scheme 4.10. The stereoselective methylation catalysed by Fom3 in fosfomycin
biosynthesis.
Scheme 4.11. Proposed approach for mechanistic studies of Sven0515-mediated
methylation using stereospecifically deuterium-labelled L-cysteine.
Scheme 4.12. Previous synthesis of stereospecifically deuterium-labelled L-cysteine at
the β-position through a deuterated thiazoline.
Scheme 4.13. Proposed method for synthesis of stereospecifically deuterium-labelled
L-cysteine at the β-position using a carbohydrate chirality template by Maeda et al.
Scheme 4.14. The synthetic route for (2R,2'R,3S,3'S)-[3,3'-2H2]cystine reported by Oba
et al.
Scheme 4.15. The mechanism of the stereospecific reduction of aldehyde 142 by
S-Alpine-Borane.
Scheme 5.1. General scheme for the biosynthesis of Streptomyces signalling molecules,
GBLs and AHFCAs, which are proposed to share a butenolide phosphate intermediate.
xv
Scheme 5.2. Synthetic route to L-cysteine with the pro-S or pro-R hydrogen atom
replaced by a deuterium atom at the β-position.
Scheme 5.3. Proposed route to characterise the absolute configuration at C-4" of
thiazostatin.
Scheme 5.4. Proposed route to characterise the absolute configuration at C-5' of
watasemycin.
xvi
List of Tables
Table 1.1. ARE sequences in the S. coelicolor methylenomycin biosynthetic gene cluster
and the S. venezuelae gaburedin biosynthetic gene cluster (W = A/T, Y = C/T).
Table 3.1. Longer and shorter DNA sequences containing each of MAREs.
Table 4.1. Proposed functions of proteins encoded by the watasemycin biosynthetic gene
cluster.
Table 4.2. Calculation of the relative amounts of the deuterium-labelled thiazostatins
from the HRMS.
Table 4.3. Calculation of the relative amounts of the deuterium-labelled watasemycins
from the HRMS.
Table 6.1. Primers and PCR conditions used in this study.
Table 6.2. Primers and conditions used for site-directed mutagenesis of MmfR.
Table 6.3. Recipe for 15% and 12% SDS-PAGE gels.
Table 6.4. Oligonucleotides for forming hairpin DNA and short annealed DNA
fragments.
Table 6.5. Recipe for 6% and 10% non-denaturing polyacrylamide gels.
Table 6.6. A typical run of an EMSA assay.
xvii
Acknowledgements
Firstly, I would like to express my deep gratitude to my supervisor, Prof. Gregory L.
Challis, for giving me the opportunity to carry out a PhD project under his patience,
guidance, support and supervision. His wide knowledge and creative way of thinking
through constructive comments brought invaluable guidance to this work. I would like
to extend my gratitude to my advisory panel members, Dr. Claudia Blindauer and Prof.
Peter J. Sadler, for their helpful discussions and advice throughout my PhD.
Secondly, I would like to thank Dr. Christophe Corre for his guidance and useful
discussions in the biology work. I am truly grateful to Dr. Lijiang Song for his kind help
and expertise in mass spectrometry, to Dr. Lona Alkhalaf for her helpful comments and
advice on my thesis. For their technical advice and helpful comments in the lab, I wish
to thank Dr. Daniel Zabala-Alvarez, Dr. Joleen Masschelein, Dr. Matthew Jenner, Dr.
Simone Kosol, Dr. Vincent Poon, Dr. Chuan Huang, Dr. Douglas Roberts, Dr. Yousef
Dashti, Dr. Mirian Rodriguez-Garcia, Dr. Emzo De Los Santos, Dr. Matias Rey, Dr.
David Withall, Dr. John D. Sidda, Daniel Griffiths, Jade Ronan, Ruby Awodi, Rebin
Salih, Gideon Idowu, Joshua Cartwright, Rakesh Saroay, Richard Gibson, Marianne
Costa, Xinyun Jian, Christian Hobson, Chris Perry, Matt Beech and all of the other
people who have made working in the Chemical Biology Research Facility such a great
experience.
I wish to thank Nicolas Malet, Peter Harrison and Dr. Yuki Inahashi for their previous
work on the project. I also wish to thank Dr. Ivan Prokes and Phil Aston for all of their
training and technical support with the NMR and mass spec aspects of the project.
I would also like to thank my friends that support me through my PhD, making these four
years memorable: Yijun Zhu, Tiantian Fu, Candace Ho, Hualong Song and many others.
Finally, I also wish to thank my parents and my brother for all of their support throughout
my PhD studies. Last but not least, I wish to thank Rubing Wang for his understanding,
encouragement and support during the time we were away from each other.
xviii
Declaration
This thesis is submitted to the University of Warwick in support of my application for
the degree of Doctor of Philosophy. It has been composed by myself and has not been
submitted in any previous application for any degree.
The work presented (including data generated and data analysis) was carried out
by the author except in the cases outlined below:
Work in Chapter 4 on chiral LC-MS was performed by Dr. Lijiang Song
Parts of this thesis have been published or accepted for publication by the author:
Zhou, S., Alkhalaf, L., de los Santos, E. L. C. and Challis, G. L., Mechanistic insights
into class B radical-S-adenosylmethionine methylases: Ubiquitous tailoring enzymes
in natural product biosynthesis. Curr. Opin. Chem. Biol. 2016, 35, 73-79.
Inahashi, Y.,† Zhou, S.,† Bibb, M. J., Song, L., Al-Bassam, M. M., Bibb, M. J. and
Challis, G. L., Watasemycin biosynthesis in Streptomyces venezuelae: thiazoline
C-methylation by a type B radical-SAM methylase. Chem. Sci. 2017, 8, accepted for
publication. († equal contribution)
xix
Abstract
Streptomyces, the largest genus of Actinobacteria, are renowned for their ability to
produce a wide variety of specialised metabolites, including many clinically used
antibiotics and other bioactive natural products. Biosynthesis of these specialised
metabolites is often tightly regulated by transcriptional regulators, some of which are
responsive to signalling molecules, e.g., 2-alkyl-4-hydroxymethylfuran-3-carboxylic
acids (AHFCAs), thus initiating a series of regulatory events.
The biosynthetic role of MmfL in AHFCA assembly has been investigated in vitro using
a synthetic substrate. MmfL has been shown to catalyse an AfsA-like reaction, involving
condensation of an ACP-bound β-ketothioester and dihydroxyacetone phosphate to form
a phosphorylated butenolide intermediate.
The interactions between DNA, AHFCA ligands and ArpA-like repressors, MmfR and
SgnR, have been investigated in vitro using electrophoretic mobility shift assays. This
work leads to a better understanding of the regulation mechanism in specialised
metabolite biosynthesis.
Heterologous expression of a putative pyochelin-like gene cluster, sven0503-sven0517
from S. venezuelae ATCC 10712, has previously been shown to result in the production
of thiazostatin and watasemycin as the main metabolic products of the cluster.
Isopyochelin, a structural isomer of pyochelin, was also identified and characterised by
comparison with synthetic standards. Sven0516 has previously been identified as the
thiazoline reductase required for the biosynthesis of all of the metabolic products of the
cluster. The class B radical SAM methylase Sven0515 was shown to be responsible for
the methylation of thiazostatin to give watasemycin. This is the first experimentally-
validated example of such a reaction in the biosynthesis of a nonribosomal peptide. From
incorporation experiments using stereospecifically deuterium-labelled cysteine, it has
been demonstrated that Sven0515-mediated methylation proceeds with abstraction of the
pro-R hydrogen atom and results in inversion of stereochemistry at the methylating
position. The absolute stereochemistry of thiazostatin and watasemycin has been
reassigned on the basis of the absolute configuration determined for isopyochelin.
xx
Abbreviations
ACP acyl carrier protein
AHFCA 2-alkyl-4-hydroxymethylfuran-3-carboxylic acid
AMP adenosine monophosphate
AMR antimicrobial resistance
ARE autoregulatory response element
ATP adenosine triphosphate
Boc t-butyloxycarbonyl
Bu butyl
CD circular dichroism
CDA calcium-dependent antibiotic
CDMT 2-chloro-4,6-dimethoxy-1,3,5-triazine
CoA coenzyme A
5′-dAdo 5′-deoxyadenosyl
dAdoH 5'-deoxyadenosine
DCC N,N'-dicyclohexylcarbodiimide
DCU dicyclohexylurea
DCM dichloromethane
DECP diethyl cyanophosphonate
DHA dihydroxyacetone
DHAP dihydroxyacetone phosphate
DIBAL-H diisobutylaluminium hydride
DMAP 4-dimethylaminopyridine
DMF dimethylformamide
DMSO dimethyl sulfoxide
DNA deoxyribonucleic acid
DTT dithiothreitol
EDC 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide
EDTA ethylenediaminetetraacetic acid
xxi
EIC extracted ion chromatogram
EMSA electrophoretic mobility shift assay
ESBL extended spectrum beta-lactamase
ESI electrospray ionization
Et ethyl
GBL γ-butyrolactone
HEP 2-hydroxyethylphosphonate
HEPES 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid
HPLC high performance liquid chromatography
HRMS high resolution mass spectrometry
HTH helix-turn-helix
HPP 2-hydroxypropylphosphonate
IPTG isopropyl β-D-1-thiogalactopyranoside
LDA lithium diisopropylamide
Me methyl
MeCbl methylcobalamin
MMF methylenomycin furan
MRSA methicillin-resistant Staphylococcus aureus
MS mass spectrometry
MS/MS tandem mass spectrometry
MT methyl transferase
MWM molecular weight marker
m/z mass/charge ratio
NAC N-acetylcysteamine
NAD(P)H nicotinamide adenine dinucleotide (phosphate)
NMM N-methylmorpholine
NMR nuclear magnetic resonance
NOESY nuclear overhauser effect spectroscopy
NRP nonribosomally biosynthesised peptide
NRPS nonribosomal peptide synthetase
xxii
OD optical density
PAGE polyacrylamide gel electrophoresis
PCD programmed cell death
PCP peptidyl carrier protein
PCR polymerase chain reaction
PLP pyridoxal phosphate
PMSF phenylmethylsulfonyl fluoride
ppm parts per million
RiPP ribosomally biosynthesised and posttranslationally modified peptide
RNA ribonucleic acid
ROS reactive oxygen species
RS radical SAM
RT room temperature
SAH S-adenosyl homocysteine
SAM S-adenosylmethionine
SCB Streptomyces coelicolor butyrolactone
SDS sodium dodecyl sulfate
SFM soya flour medium
SMM supplemented minimal medium
TBAF tetrabutylammonium fluoride
TBAI tetrabutylammonium iodide
TBDMS tert-butyldimethylsilyl
TBDMSOTf tert-butyldimethylsilyl triflate
TE thioesterase
TEMED N,N,N,N-tetramethylethylenediamine
THF tetrahydrofuran
TLC thin layer chromatography
TMS tetramethylsilane
UV ultraviolet
VB virginiae butanolide
xxiii
VRE vancomycin-resistant Enterococcus
VRSA vancomycin-resistant Staphylococcus aureus
WT wild type
Xre xenobiotic response element
YD yeast-extract dextrose
Chapter 1 : Introduction
1
1. Introduction
Chapter 1 : Introduction
2
1.1 Streptomyces
Streptomyces is the largest genus of Actinobacteria and over 600 species of Streptomyces
bacteria have been described while the number is still increasing every year.1-2 Found
predominantly in soil and decaying vegetation, Streptomyces is widely spread around the
world. As with the other Actinobacteria, streptomycetes are Gram-positive bacteria with
high GC content in their genome. Streptomyces is renowned for their ability to produce
antibiotics and other bioactive natural products with a wide range of applications in
medicine and agriculture. The production of antibiotics, as well as other specialised
metabolites, is temporally correlated to the onset of development in the Streptomyces life
cycle (Figure 1.1).3
Figure 1.1. The Streptomyces life cycle.
The life cycle of the multicellular mycelial Streptomyces starts with the germination of a
spore that grows out to form vegetative hyphae, after which a process of hyphal growth
and branching results in an intricately branched vegetative mycelium. Under adverse
conditions, such as nutrient depletion, the vegetative mycelium differentiates to form
erected sporogenic structures called aerial hyphae. This is also the moment in the life
cycle when most antibiotics are produced. The reason underlying the production of
antibiotics at that stage may be due to the need to defend the colony when it is undergoing
programmed cell death (PCD).4 When nutrient depletion occurs, the vegetative or
substrate mycelium is autolytically degraded by a PCD-like mechanism so as to provide
nutrients for the next round of biomass formation. The resulting nutrients, including
amino acids, aminosugars, nucleotides, and lipids, around the lysing substrate mycelium
Chapter 1 : Introduction
3
would inevitably attract motile competing microbes in the habitat. Thus it has been
assumed that antibiotics are produced at this time to protect the pool of nutrients.
Genetic studies of Streptomyces development regulation have revealed a collection of
developmental genes based on analysis of their mutant strains defective in different
stages of development.4 Genes that are required for the formation of aerial hyphae are
referred to as bld genes, in reference to the bald phenotype of mutants that lack the fluffy
aerial hyphae due to their defect in aerial growth. Genes that are required for the
formation of mature grey spores on the tips of the white aerial mycelium are called whi
(white) genes, corresponding to those mutants that are defective in producing the grey
spore pigment. Examples of bld genes include bldD, a highly pleiotropic transcription
factor that controls hundreds of development-related genes, and bldM, which has been
identified to play a critical role in Streptomyces differentiation, and many others.4-6
The history of antibiotics derived from the genus Streptomyces begins with the discovery
of streptothricin 1 from Streptomyces lavendulae in 1942.7 Streptothricins are broad
antimicrobial agents with a potent inhibitory activity for prokaryotic protein synthesis.
However, their cytotoxicity prevents further clinical or veterinary use. It was
streptomycin 2 from Streptomyces griseus, which came 2 years later, that triggered
systematic screening of antibiotics within this genus.8 Streptomycin is a clinically-used
aminoglycoside antibiotic typically used in the treatment of tuberculosis, always in
combination with other antituberculosis agents, and sensitive Gram-negative infections.
Its acts on the ribosomes by interfering with the binding of aminoacyl-tRNA, causing an
erroneous reading of the genetic code, inhibition of initiation of translation of mRNA
and aberrant proofreading.9-11 After the discovery of streptomycin, many other bioactive
natural products have been isolated from Streptomyces species, including antibacterials
(e.g., daptomycin 3 and chloramphenicol 4), antitumor agents (e.g., doxorubicin 5),
immunosuppressants (e.g., tacrolimus 6 and rapamycin 7), anthelmintics and insecticides
(e.g., avermectin B1a 8 and B1b 9), herbicides (e.g., phosphinothricin 10) and antifungal
agents (e.g., amphotericin B 11) (Figure 1.2).1,12 It is now estimated that over 70% of
antibiotics currently used clinically are sourced from Streptomyces species.
Chapter 1 : Introduction
4
Figure 1.2. Examples of bioactive metabolites that are produced by Streptomyces species.
The activity and producing organism for each metabolite is shown below its name.
Despite the success of the discovery of antibiotics, and advances in the process of their
production, there is an urgent need to discover and develop new antibiotics due to the
emergence of antimicrobial resistance (AMR) in microorganisms.13-14 Rising drug
resistance means infectious diseases are increasingly difficult to treat, requiring
alternative medications or higher doses, which may be more costly or more toxic. It has
been reported that a few infections are now completely untreatable due to AMR, such as
gonorrhoea.15 Some well-known examples of drug-resistant bacteria include methicillin-
resistant Staphylococcus aureus (MRSA), vancomycin-resistant S. aureus (VRSA),
Chapter 1 : Introduction
5
extended spectrum beta-lactamase (ESBL) and vancomycin-resistant Enterococcus
(VRE). However, compared to the “golden age” of antibiotic discovery, the number of
antimicrobial compounds reported per year have substantially declined.16-17 This decline
may be attributed to the reduced probability of finding novel compounds in traditional
screening and reduced funding for screening efforts. As the largest antibiotic-producing
genus, Streptomyces species play a very important role in the fight against emerging
multidrug-resistant pathogens. Several species of Streptomyces have been particularly
well studied, such as Streptomyces coelicolor A3(2), the most widely used in genetic
studies.
1.1.1 Streptomyces coelicolor A3(2)
S. coelicolor A3(2) is a model organism that has been used for more than five decades to
study the genetic and biochemical basis for the production of bioactive metabolites.18
The sequencing of the S. coelicolor A3(2) genome, as well as its two plasmids, was
completed between 2002-2004.19-21 Its linear chromosome contains 8.7 Mb with 7,825
predicted open reading frames while the giant linear plasmid SCP1 and the small circular
plasmid SCP2 is 365 kb and 31 kb, respectively. Before the DNA sequencing of
S. coelicolor A3(2), several specialised metabolites had already been discovered
(Figure 1.3),19 including actinorhodins (e.g., 12), prodiginines (e.g., undecylprodigiosin
13), calcium-dependent antibiotics (CDAs, e.g., CDA 2a 14), grey spore pigment,22
S. coelicolor γ-butyrolactones (SCBs, e.g., SCB1 15) and methylenomycins (e.g.,
methylenomycin A 16),23-24 the gene cluster of which is carried on the plasmid SCP1.
After the genome sequence of S. coelicolor was published, it became apparent that the
true potential of this species as a producer of natural products had actually been
underestimated.19 Exploitation of the genome sequence has led to the identification of
over 20 biosynthetic gene clusters, including 18 cryptic ones, for specialised metabolites.
The products of the cryptic gene clusters were previously unknown, with many not
expressed under standard laboratory conditions. A range of metabolic products have been
confirmed to be produced from S. coelicolor A3(2),18 such as desferrioxamines (e.g.,
desferrioxamine E 17), coelichelin 18, geosmin 19, albaflavenone 20,
2-methylisoborneol 21, carotenoids (e.g., isorenieratene 22), hopanoids (e.g.,
aminotrihydroxybacteriohopane 23), germicidins (e.g., germicidin A 24), flaviolin
oligomers (e.g., flaviolin dimer 25), methylenomycin furans (e.g., AHFCA1 26) and
Chapter 1 : Introduction
6
coelimycins (e.g., coelimycin P1 27) (Figure 1.3). Among all of the above metabolic
products, at least six of them, including methylenomycins,25 actinorhodins,26
prodiginines,27 CDAs,28 coelimycins,18,29-30 and albaflavenone,31-32 have shown
antibiotic activity while desferrioxamines and coelichelin are known to function as
siderophores in S. coelicolor.33-35
Figure 1.3. Representative structures of specialised metabolites produced by S.
coelicolor.
Therefore, it appears that the potential of S. coelicolor for specialised metabolite
production is much greater than originally anticipated from classical bioassay-guided
discovery approaches. Indeed, this does not only occur in S. coelicolor, but also in other
species, like S. griseus,36 Streptomyces avermitilis and Streptomyces venezuelae,37-38
Chapter 1 : Introduction
7
which are all predicted to contain more than 30 natural product biosynthetic gene clusters
from analysis of their genome sequences. This predicted number of putative natural
products is much higher than has been identified so far. This has prompted extensive
research into those cryptic, silent, or sleeping metabolic products and methods for
activating their biosynthesis.39-42
1.2 Streptomyces signalling molecules
1.2.1 Structural types of Streptomyces signalling molecules
Of particular interest to many specialised metabolite biosynthetic pathways in bacteria
are involvement of diffusible and low molecular weight signalling molecules, also known
as ‘bacterial hormones’ or autoinducers.43 These molecules control complex regulatory
systems in combination with transcriptional regulators, activators or repressors, in a
pathway-specific or pleiotropic manner. To date there are two main classes of signalling
molecules found in Streptomyces species, the γ-butyrolactone (GBL)-type molecules and
the furan-type autoinducers, 2-alkyl-4-hydroxymethylfuran-3-carboxylic acids
(AHFCAs), which were previously known as methylenomycin furans (MMFs).44-45
Repressors which recognise these types of signalling molecules all belong to the TetR
subfamily of ArpA-like transcriptional repressors and are often involved in regulation of
antibiotic biosynthesis. These transcriptional repressors and their regulation involving
small signalling molecules are discussed in Section 1.3.
The first example of γ-butyrolactone signalling molecules is autoregulatory factor
(A-factor) 28 (Figure 1.5) which was isolated from Streptomyces species in 1967.46 It has
been found to induce morphological differentiation, such as aerial hyphae formation, and
production of specialised metabolites, including streptomycin and a yellow pigment
grixazone, in S. griseus.47-48 Thus, A-factor has been established to have a pleiotropic
effect.
Since the discovery of A-factor, other homologues with a γ-butyrolactone have been
identified as signalling molecules for specialised metabolism in a number of
Streptomyces species and related genera.49 This has thus opened a novel paradigm of the
characteristic regulatory system for cellular differentiation in prokaryotes. Based on the
currently elucidated structure of GBLs, these molecules can be generally divided into
three groups (Figure 1.4) according to the substituent and its stereochemistry at C-6: i)
Chapter 1 : Introduction
8
6-keto type, such as A-factor 28, which is the only natural example of the 6-keto type of
GBLs reported. A-factor readily undergoes epimerisation due to the relative acidity of
the proton at C-2 position; ii) IM-2 type, possessing a 6-β-hydroxyl group; iii) VB type,
possessing a 6-α-hydroxyl group. Both IM-2 and VB type GBLs belong to the 6-hydroxy
type. It has been shown that the absolute stereochemistry of carbons 2, 3 and 6, as well
as the length of the alkyl side chains, confers the specificity and affinity of each
γ-butyrolactone receptor for its cognate ligand(s).50-51
Figure 1.4. Three structural types of GBLs.
There are a range of GBLs isolated and characterised from Streptomyces species. They
include: factor I 29 from Streptomyces viridochromogenes,52-53 Gräfe factors 30-32 from
Streptomyces bikinensis and Streptomyces cyaneofuscatus;53-54 virginiae butanolides
(VB) A-E 33-37 isolated from Streptomyces virginiae;53,55-56 IM-2 38 from Streptomyces
lavendulae FRI-5 and SCBs 1-3 15, 39, 40 from S. coelicolor for regulation of both
actinorhodin and undecylprodigiosin production.50,57 Their structures are shown in
Figure 1.5 although the absolute stereochemistry of factor I 29 and Gräfe factors 30-32
has not been clearly elucidated.
Chapter 1 : Introduction
9
Figure 1.5. Structures of Streptomyces GBL-type signalling molecules. The absolute
stereochemistry of factor I and Gräfe factors are not clear.
From S. coelicolor, the furan type of signalling molecules, AHFCAs 1-5 26, 41-44
(Figure 1.6) have been identified and shown to control methylenomycin biosynthesis.45,58
AHFCA5-7 44-46 have also been characterised and proposed to function as signalling
molecules for regulation of gaburedin biosynthesis in S. venezuelae ATCC10712.59-60
In addition, another signalling molecule avenolide 47 (Figure 1.6), has been
demonstrated to regulate avermectins (e.g., 8 and 9 in Figure 1.2) production in S.
avermitilis.61 After its discovery, similar butenolides SRB1 48 and SRB2 49 (Figure 1.6)
have been reported to induce the production of lankacidin and lankamycin in
Streptomyces rochei.62 It seems that they represent a new group of butenolide-type
signalling molecules.
Chapter 1 : Introduction
10
Figure 1.6. Structures of Streptomyces AHFCA- and butenolide-type signalling
molecules. SRB1 and SRB2 exist as mixtures of interconverting epimers at C-4.62
1.2.2 Biosynthesis of γ-butyrolactones
As the most well-studied group of signalling molecules in Streptomyces, the biosynthesis
of GBLs has attracted a great deal of interests since the discovery of the first
γ-butyrolactone molecule A-factor. Investigation into A-factor biosynthesis was first
conducted by Horinouchi and co-workers as early as 1984. A-factor production was
restored by insertion of the afsA gene into an A-factor-deficient S. griseus mutant,
indicating that AfsA may be an important enzyme required for A-factor biosynthesis.63-
65
Prior to identification of the specific enzymes responsible for production of GBLs, the
biosynthesis of VB-A 33 was investigated by incorporation experiments using stable
isotope-labelled precursors in Streptomyces antibioticus.66 The feeding experiments
revealed that VB-A is formed from two molecules of acetate and one molecule of both
glycerol and isovalerate (Figure 1.7). It was proposed that β-keto acyl CoA is the key
precursor, which couples to a C3 unit that is oxidised from the glycerol molecule.
Chapter 1 : Introduction
11
Figure 1.7. Incorporation results using 13C-labelled glycerol, acetate and isovalerate for
investigation of VB-A 33 biosynthesis by Sakuda and co-workers.66
The biosynthesis of GBLs didn’t become much clearer until in vitro reconstitution of
AfsA by Kato and co-workers in 2007.67 AfsA has been demonstrated to be the key
enzyme required for A-factor biosynthesis, catalysing β-ketoacyl transfer from an acyl
carrier protein (ACP)-bound 8-methyl-3-oxononanoyl thioester 51 to the hydroxyl group
of dihydroxyacetone phosphate (DHAP) 50 to form an 8-methyl-3-oxononanoyl-DHAP
ester intermediate 52 (Scheme 1.1). After formation of the phosphorylated ester
intermediate 52, three further reactions are required for A-factor biosynthesis, including
an intramolecular aldol reaction that leads to the butenolide structure 55, a reduction step
for the C2-C3 double bond in the butenolide intermediate and a dephosphorylation step
to give the final A-factor 28. It was also proposed that the three subsequent steps may
occur in a different order, i.e. dephosphorylation of the ester intermediate 52 is followed
by the aldol condensation and final reduction of the butenolide 54 to A-factor (Scheme
1.1).
Chapter 1 : Introduction
12
Scheme 1.1. Biosynthetic route to A-factor as proposed by Kato and co-workers.
A range of C3 compounds were first tested with His-AfsA, leading to identification of
DHAP 50 as the C3 substrate for the AfsA-catalysed reaction. Then the enzymatic assay
was undertaken using recombinant His-AfsA, 32P-labelled DHAP and the synthetic 8-
methyl-3-oxononanoyl-N-acetylcysteamine (NAC), which mimics the corresponding β-
ketoacyl-ACP. The reaction resulted in a product with radioactivity on radio-thin layer
chromatography (TLC), which was demonstrated to be the ester intermediate, 8-methyl-
3-oxononanoyl-DHAP ester 52, by comparing to its synthetic standard. The butenolide
is then formed by a nonenzymatically catalysed intramolecular aldol reaction of the ester
intermediate 52.
Chapter 1 : Introduction
13
BprA (butenolide phosphate reductase), an NADPH-dependent reductase that is encoded
by a gene directly downstream of afsA, was confirmed in vitro to be responsible for
reduction of the butenolide 55 to form the butanolide 56. However, it was shown that the
phosphatases and reductases commonly present in bacteria are able to catalyse similar
reactions to afford A-factor. For example, introduction of afsA into E. coli caused the
host to produce A-factor analogues, with straight side chains, due to the differences in
fatty acid metabolism between Streptomyces species and E. coli. Streptomyces species
can produce branched fatty acids from starter units such as isobutyryl-CoA and
methylbutyryl-CoA from amino acid degradation, whereas E. coli predominantly uses
acetyl CoA and thus does not produce branched fatty acids.68-69
An AfsA homologue, ScbA in S. coelicolor, has been proposed for the biosynthesis of
SCBs.70-71 Takano and co-workers conducted in silico analysis of the AfsA family of
proteins, including ScbA, and showed that they have similarity to the fatty acid synthesis
enzymes FabA and FabZ. Meanwhile, mutation of two predicted active sites in ScbA led
to abolishment of the production of SCBs in S. coelicolor. Other AfsA homologues have
also been found in corresponding γ-butyrolactone producers, e.g. BarX in S. virginiae
and FarX in S. lavendulae.72-73 It is reasonable to assume that these homologues have
similar activity as AfsA in γ-butyrolactone biosynthesis. It was also suggested by Kato
and co-workers that a BprA orthologue, ScbB, which has 76% identity in amino acid
sequence to BprA, catalyses the reduction of C2-C3 double bond to form the butanolide
phosphate 56 in SCB biosynthesis.67
Both SCBs in S. coelicolor A3(2) and VBs in S. virginiae contain a hydroxyl group at
C-6 instead of the keto group in A-factor, indicating the presence of additional reductases
for this reduction. Due to the absence of such reductase in S. griseus, C-6 of A-factor
remains as a keto group. One such enzyme, BarS1 in S. virginiae, has been characterised.
BarS1 was isolated as an NADPH-dependent reductase that reduces the 6-oxo group of
the penultimate intermediate in the VB biosynthesis.74 A BarS1 orthologue, SCO6264,
has been proposed to be involved in SCB biosynthesis for a similar reduction reaction in
S. coelicolor.70 More investigations are needed to experimentally determine the specific
functions of these enzymes.
Chapter 1 : Introduction
14
1.2.3 Biosynthesis of AHFCAs
The furan-type signalling molecules, AHFCAs 1-5, were demonstrated to regulate
methylenomycin biosynthesis in S. coelicolor by Corre and co-workers in 2008.45 A
putative operon of three genes within the mmy cluster, mmfLHP, was proposed to direct
the biosynthesis of these AHFCA signalling compounds. Deletion of the three genes
caused the loss of methylenomycin production.58 Comparative metabolic profiling has
led to the identification and structure elucidation of the AHFCA family members,
AHFCAs 1-5 26, 41-44 (Figure 1.6), as the products of the mmfLHP genes.45
The AfsA homologue, MmfL, has been proposed to play a key role in AHFCA
biosynthesis. Methylenomycin production can be restored in an mmfL-deficient strain by
addition of an EtOAc extract of the culture supernatant from a strain containing mmfL.
Meanwhile, expression of the mmfLHP putative operon in S. coelicolor M512, which
lacks the SCP1 plasmid, led to the production of the five natural AHFCAs 1-5.
Figure 1.8. Alignment of the amino acid sequences of AfsA from S. griseus and the S.
coelicolor AfsA homologues ScbA and MmfL. Identical and conservatively substituted
amino acids are highlighted with asterisk and dots, respectively.
Chapter 1 : Introduction
15
MmfL shares 27% identity and 43% similarity in amino acid sequence to AfsA.75
Sequence alignment of MmfL with AfsA in S. griseus and ScbA in S. coelicolor, both of
which are key enzymes involved in the biosynthesis of GBL-type signalling molecules,
are shown in Figure 1.8.
It was proposed that MmfL catalyses an analogous reaction to AfsA, involving
condensation of the β-ketoacyl thioesters derived from fatty acid metabolism with DHAP
to form phosphorylated butenolide intermediates. It was also hypothesised that
subsequent dephosphorylation, catalysed by MmfP, and further conversion catalysed by
MmfH, yield the AHFCA molecules from the phosphorylated butenolide intermediates
(Scheme 1.2).
Scheme 1.2. Proposed pathway for the biosynthesis of AHFCAs, taking AHFCA1 as an
example.75
During the initial elucidation of the structures of AHFCAs, preliminary evidence has
been provided for this hypothesis. Incorporation experiments using deuterium-labelled
precursors for starter units of fatty acid biosynthesis showed that different fatty acid
starter units are incorporated in each of the AHFCAs, yielding the different side chains.
One example from the feeding experiments is shown in Figure 1.9.
Chapter 1 : Introduction
16
Figure 1.9. An example of the feeding experiments using deuterium-labelled precursors
to yield the AHFCAs.
Davis and co-workers proposed that the divergence between GBL and AHFCA
biosynthesis may arise due to the orientation of DHAP relative to the β-ketothioester
precursors in the reactions catalysed by different AfsA-like enzymes.76 This hypothesis
implied that AHFCA biosynthesis would not proceed via a butenolide intermediate.
However, subsequent incorporation experiments using 13C-labelled glycerols labelled at
the pro-R and pro-S carbon atoms confirmed the orientation of DHAP with respect to the
β-ketothioester, thus supporting that AHFCA biosynthesis proceeds via a butenolide
intermediate.75
More recently, the three enzymes, MmfL, MmfH and MmfP, were studied in vitro to
establish their specific roles in the assembly of AHFCAs (PhD thesis by N. Malet).77
Each enzyme was overproduced in E. coli and purified as recombinant His6-tagged
protein. MmfL was shown to catalyse the condensation of DHAP 50 and the synthetic
NAC β-ketothioester 61 to give a putative phosphorylated butenolide product 59
(Scheme 1.3), which could not be detected due to its high polarity. Treatment of the
product of the MmfL-catalysed reaction with alkaline phosphatase yielded a
dephosphorylated product. However, experimental evidence is needed for structure
confirmation of this dephosphorylated product as the butenolide 60. MmfP has been
demonstrated to be a Mg2+ dependent phosphatase capable of catalysing the
dephosphorylation of the presumed phosphorylated butenolide 59 of the MmfL-catalysed
reaction to yield butenolide 60 in in vitro enzymatic assays. Interestingly, FAD was
shown to be capable of transforming the butenolide intermediate 60 into an aldehyde
analogue 62 of AHFCA1, which could be reduced using sodium borohydride to yield
AHFCA1 26 (Scheme 1.3).77 However, more work is needed to ascertain the function of
each enzyme in AHFCA biosynthesis.
Chapter 1 : Introduction
17
Scheme 1.3. Summary of the in vitro enzymatic studies on AHFCA biosynthesis by N.
Malet.77
1.3 Transcriptional regulation of specialised metabolites in Streptomyces
1.3.1 TetR transcriptional repressors
Signalling molecules, such as GBLs and AHFCAs, are involved in regulation of
corresponding specialised metabolite biosynthesis upon recognition and interaction with
their cognate receptor proteins. Their receptors are known as the ArpA-like
transcriptional repressor proteins. The ArpA class of the TetR family is an example of a
transcriptional repressor involved in regulation of specialised metabolite biosynthesis.
Such repressors are widespread in Streptomyces bacteria.
The TetR family of regulators act as repressors in transcription and belong to the one-
component system in which a single polypeptide contains both a sensory domain and a
DNA-binding domain.78 This family is named after the member of this group that has
been most completely characterised genetically and biochemically, the tetracycline
resistance (TetR) protein, which was first discovered in E. coli.79 TetR proteins generally
function as homodimers, with each subunit consisting of a ligand-binding domain and a
highly conserved helix-turn-helix (HTH) DNA-binding domain at the N-terminus.
Members of this family are best known for their roles as regulators of antibiotic efflux
pumps, which contribute significantly to antibiotic self-resistance mechanisms. Apart
from this, they are also known to regulate numerous aspects of bacterial physiology,
Chapter 1 : Introduction
18
including metabolism of amino acids, lipids, and other carbon and nitrogen-containing
compounds.
TetR tightly regulates expression of tetA by binding specifically to the operator located
in the tetR-tetA intergenic region (Figure 1.10). Tetracycline 63 binding to the repressor
initiates conformational changes in TetR, thus reducing the affinity of its DNA binding
domain for the tetA promoter within the tetA-tetR intergenic region. Dissociation of TetR
from the DNA allows RNA polymerase to bind to the tetA promoter, resulting in
transcription of tetA. The translated TetA protein functions as efflux pump for
tetracycline, thus conferring self-resistance to tetracycline. Since TetR repressors are
homodimeric, they often interact with semi-palindromic or dyad sequences.80 The HTH
DNA-binding domain binds to major grooves of DNA.
Figure 1.10. Mode of action of TetR. TetR repressor binds to the tetA promoter and
represses tetA expression. The binding of tetracycline to TetR releases it from the
promoter, thus allowing transcription of tetA. The TetA protein exports tetracycline out
of the cell.
1.3.2 ArpA-like transcriptional repressors
As a cognate receptor for A-factor, which was discovered from S. griseus in the 1960s,
the A-factor receptor protein (ArpA) from S. griseus was the first and most extensively
studied ArpA-family repressor in Streptomyces bacteria. The ArpA class of proteins
belongs to the TetR family and act as transcriptional repressors for regulation of
specialised metabolites biosynthesis.
The ArpA-like repressors are widespread in Streptomyces bacteria. Similar to the
tetracycline/TetR system, when the intracellular concentration of A-factor reaches a
critical point, it binds to ArpA and causes conformational change in the protein, thus
leading to release of ArpA from its cognate operator sequence, which is known as an
Chapter 1 : Introduction
19
autoregulatory response element (ARE) located in the promoter of adpA. Therefore, it
results in the transcription and translation of adpA. The expressed AdpA is a pleiotropic
transcriptional activator implicated in the regulation of specialised metabolites
production and morphological differentiation (Figure 1.11). In the streptomycin
biosynthetic gene cluster of S. griseus, StrR functions as a pathway-specific
transcriptional activator required to transcribe other genes within the cluster.81 AdpA was
identified to bind the operator region of strR and activate its expression, which leads to
induction of streptomycin production.82 Meanwhile, AdpA indirectly activates
expression of griR, a pathway-specific activator gene that regulates biosynthesis of the
yellow pigment grixazone.83 It is also required for expression of other members of the
adpA regulon, some of which are required for morphological differentiation.84
Figure 1.11. The A-Factor/ArpA regulatory cascade in S. griseus leading to transcription
of adpA and subsequent morphological development and production of specialised
metabolites, such as streptomycin and grixazone.
Analogous A-factor/ArpA systems in Streptomyces bacteria have been identified by
genome mining, such as the AHFCA/SgnR/GbnR system regulating gaburedin
biosynthesis in S. venezuelae,59 the AHFCA/MmfR/MmyR system regulating
methylenomycin biosynthesis in S. coelicolor,45,58 and the SCB/ScbR/ScbR2 system
controlling the biosynthesis of coelimycins in S. coelicolor.71,85
Chapter 1 : Introduction
20
1.3.3 Regulation of methylenomycin biosynthesis in S. coelicolor
Methylenomycin A 16 (Figure 1.3), an unusual epoxycyclopentenone antibiotic, is one
of the antibiotics produced by S. coelicolor A3(2) and its biosynthetic gene cluster,
containing a total of 21 genes, are located on the 365 kb linear plasmid SCP1.
Organisation of the methylenomycin gene cluster, including biosynthetic, resistance and
regulatory genes, in S. coelicolor is shown in Figure 1.12. Methylenomycin was also
reported to be produced by Streptomyces violaceoruber SANK 95570,86 in which its gene
cluster is more than 99% identical at DNA level to that in S. coelicolor.20
Figure 1.12. Organisation of the methylenomycin biosynthetic gene cluster in S.
coelicolor A3(2).
It was noted that there are genes encoding for two repressors, mmfR and mmyR, in the
methylenomycin (mmy) gene cluster in S. coelicolor. The products of both of them show
sequence similarity to ArpA except that MmyR lacks two conserved residues that are
present in ArpA that probably play a role in DNA- and ligand-binding. Deletion of these
two repressors gives distinct phenotypes. Deletion of mmyR resulted in overproduction
of methylenomycin, whereas deletion of mmfR had the opposite effect, causing
substantial loss of production. However, the presence of the mmfR gene in the absence
of mmfLHP led to greatly reduced methylenomycin production. Further studies showed
that methylenomycin production was restored in the mmfR-containing strain when
supplied with AHFCAs, whereas this was not observed in the mmyR-carrying strain. This
indicates that MmfR, not MmyR, is sensitive to AHFCAs.58 Thus, the cassette of five
genes, mmfR, mmfL, mmfH, mmfP and mmyR, forms the AHFCA-dependent signalling
system involved in the regulation of methylenomycin biosynthesis.
Similarly to the A-factor/ArpA regulation system, AREs are identified in the mmy gene
cluster and known as MAREs, corresponding to a general consensus proposed by Folcher
and co-workers as shown in Table 1.1.87 These sequences are highly conserved short (18
Chapter 1 : Introduction
21
bp) and almost palindromic DNA fragments, located in the intergenic regions of mmfL-
mmfR and mmyY-mmyB as well as the region upstream of mmyR.
Table 1.1. ARE sequences in the S. coelicolor methylenomycin biosynthetic gene cluster
and the S. venezuelae gaburedin biosynthetic gene cluster (W = A/T, Y = C/T).
Name Intergenic region Sequence (18 bp)
ARE consensus AWACNNACYNNNCGGTTT
MARE1 mmfR-mmfL ATACCTTCCCGCAGGTAT
MARE2 mmyY-mmyB AAACCTTCGGGAAGGTTT
MARE3 upstream of mmyR ATACCTTCCCGAGGGTAT
ARE1 sgnR-sgnL ATACATTCTCGAAGGTAT
ARE2 upstream of gbnA ATACCTTCGTGAAGGTTT
ARE3 gbnR-sgnH ATACCTTCCAGACCGAAT
Binding of the repressor protein(s) to the MAREs is proposed to prevent the transcription
of the adjacent genes. Previous electrophoretic mobility shift assays (EMSAs) have
shown that MmfR is able to bind the mmfL-mmfR intergenic DNA sequence containing
MARE1 and can be released by the addition of AHFCAs.77 Recently, the crystal structure
of MmfR with AHFCA2 bound has been solved by the Corre and Challis groups (Figure
1.13). This provides a clear picture of the mode of interaction of MmfR with AHFCA
ligands. MmyR, which is 35% identical and 56% similar to MmfR in amino acid
sequence, is also proposed to bind to the same ARE sequences. However, in contrast to
MmfR, preliminary results showed that MmyR is not responsive to the same AHFCA
signalling molecules.
Figure 1.13. X-ray crystal structure of MmfR with AHFCA2 bound.
Chapter 1 : Introduction
22
The mmyB gene encodes a protein with a likely DNA-binding motif related to that of the
xenobiotic response element (Xre) family of transcriptional regulators. The in-frame
deletion of mmyB results in the loss of methylenomycin production, suggesting that
MmyB is a transcriptional activator for the biosynthesis of methylenomycin. It has been
proposed that MmyB may be the direct target of the putative AHFCA/MmfR/MmyR
regulation system.58 From bioinformatics analyses of the mmy gene cluster, there are
three short sequences (known as ‘B-boxes’) proposed to be the binding sites for MmyB,
two of which are located in the operator regions of mmyB-mmyY and one situated
upstream of mmyT.58 These sequences are distinct from the ARE sequences. The putative
function of MmyB is binding to the “B-box” sequences and activating transcription of
the structural genes in the mmy gene cluster, including the putative mmyTOG,
mmyBQEDXCAPK and mmyYF operons, leading to methylenomycin production. It was
noted that both mmfL and mmyB genes contain the TTA codon, accounting for the
dependence of methylenomycin production on bldA, which encodes the only tRNA for
the rare UUA leucine codon. Therefore, BldA is required for translation of mmfL and
mmyB. BldA is a global regulator of morphological differentiation and metabolite
production, which predominantly accumulates during late exponential and stationary
phase.88
A regulatory cascade has been proposed for the regulation of methylenomycin
biosynthesis in S. coelicolor (Figure 1.14). Transcription of mmyB, mmfL, mmyR and
itself is tightly regulated in the presence of MmfR repressor by binding to AREs. In
presence of mature BldA, mmfL is translated to a functional protein and assembles the
AHFCA signalling molecules. When the AHFCAs are produced to reach a critical
concentration, they bind to the MmfR repressor and cause derepression of MmfR binding
to the operator within the mmfL-mmfR intergenic region, leading to a significant increase
of the production of AHFCAs. The increased concentration of AHFCAs leads to
derepression of MmfR on the mmyB promoter and thus the transcriptional activator
MmyB can be produced. MmyB then binds to its operator regions (‘B-boxes’), initiating
the transcription of the structural genes to produce methylenomycins. MmyR repressor
is proposed to act as an off-switch to stop the production of methylenomycin compounds.
Chapter 1 : Introduction
23
Figure 1.14. Proposed signalling cascade with MmfR involved in methylenomycin
biosynthesis. MARE sequences for MmfR binding are located in the intergenic regions
of mmfL-mmfR, mmyY-mmyB and the region upstream of mmyR. MmyR is also proposed
to bind the same MAREs.
Analogous AHFCA-dependent signalling gene clusters have been identified from other
Streptomyces bacteria, such as in S. venezuelae.59 Sequence analysis of the gaburedin
(gbn) gene cluster revealed the presence of the mmyR homologue gbnR, as well as the
mmfR homologue sgnR and the mmfLPH homologues sgnLPH. Three highly conserved
ARE sequences identified from the gbn cluster are almost identical in sequence to the
ARE consensus from the mmy cluster and located in similar positions as those MAREs
present in the mmy gene cluster, including the region upstream of gbnA, the intergenic
regions of sgnR-sgnL and gbnR-sgnH as shown in Figure 1.15. These sequences are
proposed to be the AREs for binding by the MmfR homologue SgnR. Genetic
inactivation of the analogous mmyR repressor, gbnR, in S. venezuelae led to the discovery
of a novel group of natural products, the gaburedins.59
Chapter 1 : Introduction
24
Figure 1.15. Location of AREs, which are shown in purple, in the S. venezuelae
gaburedin biosynthetic gene cluster.
1.4 Siderophores
Siderophores are low-molecular-weight (< 10 kDa) iron-chelating compounds produced
by grasses and microorganisms such as bacteria and fungi.89-90 Their name is from the
ancient Greek words sidero ‘iron’ and phore ‘carriers’ meaning ‘iron carriers’. Iron is an
essential element that plays an important role in a diverse array of biological processes
for eukaryotes and prokaryotes. The indispensable nature of iron is due to its involvement
in several key metabolic processes including amino acid synthesis, tricarboxylic acid
cycle activity, DNA replication, cellular respiration and electron transport.91 Due to the
interconversion between the Fe3+ (ferric) and Fe2+ (ferrous) redox states, it is able to
participate in single electron transfers and mediate common redox reactions as a versatile
redox-active catalyst. However, free bioavailable iron in solution is extremely limited
(10−8-10−9 M) and below the range that supports microbial growth (∼10−6 M).92 This is
due to its very low solubility under physiological conditions and consequent formation
of ferric oxyhydroxide precipitates, though iron is actually one of the most abundant
elements on Earth.
In response to iron limitation in their environment, many bacteria can produce different
structural types of siderophores and excrete them into the extracellular environment,
where the siderophore acts to sequester and solubilise the iron. The problem of iron
deficiency is also sometimes faced by plants,93 especially for those grown in calcareous
soils with high pH, which decrease the affinity of plants for iron and thus impede iron
uptake mechanism. A great variation can be seen in siderophore structures produced by
bacteria, accompanied by different levels of iron chelating affinity. Siderophores usually
Chapter 1 : Introduction
25
form a stable complex preferentially with Fe3+ compared to other naturally occurring
abundant metal ions. Based on the co-ordinating groups that chelate the ferric iron, they
are generally classified into several groups, including the most common catecholates,
hydroxamates, carboxylates, and some containing distinct binding groups such as
salicylic acid, oxazoline or thiazoline nitrogen, or having a mix of these chemically
different classes of co-ordinating groups such as pyoverdins.89
The catecholate type of siderophore is characterised by the presence of phenolate or 2,3-
dihydroxy benzoate binding groups. Examples include the three catecholate siderophores
produced by Azotobacter vinelandii in iron-limited medium: the monocatecholate
aminochelin 64, the dicatecholate azotochelin 65 and the tricatecholate protochelin 66
(Figure 1.16).94-97 A typical triscatechol siderophore is enterobactin 67, a cyclic trimer
composed of 2,3-dihydroxy-N-benzoylserine produced by bacteria of the family
Enterobacteriaceae, including Escherichia coli and many others such as Aerobacter
aerogenes and Erwinia herbicola.98-99
Figure 1.16. Representative examples of microbial siderophores.
Hydroxamate siderophores, produced mainly by fungi as well as some Streptomyces and
Pseudomonas species, are mostly trihydroxamates, or dihydroxamates, leading to
Chapter 1 : Introduction
26
formation of hexadentate, tetradentate and bidendate ligands. This type of siderophores
include the trihydroxamate ferrichromes and the dihydroxamate aerobactin 68, as well
as the desferrioxamines mentioned previously (e.g., desferrioxamine E 17 in Figure 1.3)
secreted by many Streptomyces species including S. coelicolor A3(2). Another type of
siderophores chelate ferric iron by their hydroxyl carboxylate and carboxylate groups
such as rhizoferrin 69, which is produced by fungi, and its enantiomer, enantio-
rhizoferrin produced in some bacteria.100 In addition, certain siderophores have mixed
ligands besides the above iron-binding groups, like mycobactin 70, which is isolated
from Mycobacteria with the presence of two hydroxamates, a phenolate and an oxazoline
nitrogen in its structure (Figure 1.16).101
1.4.1 Bioactivity of siderophores
Microbial siderophores are not only important in microbial physiology but also have
wide applications in various fields, such as agriculture to improve soil fertility and
biocontrol, environmental applications and medicinal applications.89 Clinically,
siderophore-based drugs can be used to efficiently treat iron overload diseases. A
siderophore from Streptomyces pilosus, desferrioxamine B, is marketed as mesylate salt
under the trade name Desferol for removing excess iron from the body in thalassemia
treatment. Remarkably, the potency of some antibiotics has been elevated by binding into
the iron-binding functional groups of siderophores, for example, antimalarial
siderophores combating iron-dependent malaria parasites, and sideromycins (e.g.,
albomycin 71 in Figure 1.16), a group of iron-chelating antibiotics produced by
Streptomyces species exhibiting good antimicrobial activity.102 Inspired by this, one
promising strategy has been developed as so-called ‘Trojan Horse’ antibiotics, which are
covalently bonded siderophore-antibiotic conjugates.103 The siderophore part of the
molecule can bind to ferric ion and be recognised by the membrane transport system,
facilitating the uptake of the antibiotic warhead into the cell.
1.4.2 Biosynthesis of siderophores
There are two main pathways for siderophore biosynthesis, one of which involves the
use of nonribosomal peptide synthetases (NRPSs) multienzymes while the other is NRPS
independent.104 NRPSs are large multi-modular enzymes that produce peptides in a
sequential fashion, without the need for an RNA template. The order in which the
constituent amino acids are incorporated into the peptide product is dictated by the order
Chapter 1 : Introduction
27
of the NRPS modules.105 An NRPS chain extension unit normally contains three domains:
the adenylation (A) domain, the peptidyl carrier protein (PCP) domain, which is also
known as the thiolation (T) domain, and the condensation (C) domain.
The A domain of each module of the NRPS recognises the substrate and catalyses the
adenylation of its carboxyl group for activation by reaction with ATP. This activated
ester is then covalently linked as its thioester on the PCP domain. It uses terminal thiol
of a post-translationally installed phosphopantetheine arm to capture the activated
carboxyl group of the adenylate. The C domain catalyses the direct transfer to another
acyl amino acid intermediate on the adjacent downstream module to form a peptide bond.
In some NRPS modules, C domain is replaced by a heterocyclisation (Cy) domain that
catalyses heterocycle formation by a reaction of β-amino thiol group in the substrate
attached to the PCP domain of the upstream module.
The growing chain is covalently bound to the PCP domain in successive modules
throughout the assembly process. There is also a thioesterase (TE) domain, which is
usually present in the final module. Hydrolysis or cyclisation results in the release of the
assembled chain from the NRPS.
1.5 2-Hydroxyphenothiazolines
As well as having antioxidant, antibacterial and antifungal activity, many
2-hydroxyphenylthiazolines have iron-chelating properties and act as siderophores
(Figure 1.17). They are therefore produced and secreted by some pathogens as virulence
factors.105-107 Members of the family include anguibactin 72, a siderophore produced by
Vibrio anguillarum 775, which is responsible for the septicemic disease vibriosis in
fish,108-109 yersiniabactin 73, a siderophore that is an important virulence factor for some
pathogenic Yersinia strains such as Yersinia enterocolitica110 and pyochelin 74, a
siderophore first isolated from the Gram-negative Pseudomonas aeruginosa, which can
opportunistically cause infections in plants and animals, including humans.111 In
Pseudomonas species, two siderophore-mediated systems are utilised for iron uptake,
one mediated by the siderophore pyochelin and the other one mediated by the fluorescent
siderophore pyoverdin.89 Another member of the 2-hydroxyphenylthiazoline family,
aerugine 75, also identified from the Pseudomonas species, showed antifungal activity
against Colletotrichum orbiculare as well as antioomycete activity against Phytophthora
capsici and Pythium ultimum.112-113
Chapter 1 : Introduction
28
Interestingly, Pseudomonas protegens (formerly Pseudomonas fluorescens) Pf-5 and
CHA0 produce the enantiomer of pyochelin, enantio-pyochelin 76 (Figure 1.17), which
has also been shown to function as a siderophore.114 Some members of the
2-hydroxyphenylthiazoline family are produced by Streptomyces species, including
pulicatins A-E 77-81,115 thiazostatin 82 and watasemycin 83 (Figure 1.17),116-117 which
are discussed in detail in Section 1.5.3. More recently, new members of this family,
including ulbactins F/G 84, have been isolated from a Brevibacillus sp. and shown to
have anticancer activities.118 Despite having a structure similar to pyochelin 74 and
enantio-pyochelin 76, it is not currently known whether these compounds are also able
to function as siderophores.
Chapter 1 : Introduction
29
Figure 1.17. Representative members of the 2-hydroxyphenylthiazoline family
compounds found in bacteria.
1.5.1 Pyochelin and enantiopyochelin
4'R, 4"R-Pyochelin 74 is a mixed-type siderophore containing a phenolate, a carboxylate,
a thiazoline and a thiazolidine nitrogen atom. It was first isolated from P. aeruginosa,111
and later found to be also produced by some members of the Burkholderia cepacia
complex,119 and plant-associated P. fluorescens,120 as well as the Gram-positive
Streptomyces scabies 87-22.121 It should be noted that pyochelin is normally present as
an equilibrium mixture of two diastereomeric forms due to isomerism at C-2". Apart
from pyochelin, this also occurs in other thiazolidine derivatives including yersiniabactin
Chapter 1 : Introduction
30
73,110 ulbactins F/G 84,118 thiazostatin 82 and watasemycin 83 (Figure 1.17).116-117 A
possible mechanism for the epimerisation was proposed by Seto et al. for thiazostatin
(Figure 1.18).116
Figure 1.18. Proposed mechanism for epimerisation at C-2 of 2,4-disubstituted
thiazolidine compounds.116
As one of the two important siderophore-mediated iron-uptake systems in Pseudomonas
species, the structure, iron-chelating mode and biosynthesis of pyochelin have been
clearly elucidated. The isolation of a 1:1:1 complex of Fe3+, pyochelin and cepaciabactin,
a bidendate ligand, suggests that four of the ligand sites of Fe3+ are satisfied by pyochelin
while the remaining two bind cepaciabactin.122 This has been further supported by the
X-ray structure analysis of the 1:1 complex of pyochelin and Fe3+, in which four of the
six octahedral coordination sites of Fe3+ are occupied by the phenolate and the
carboxylate oxygen and by the two nitrogen atoms of the two heterocycle rings while the
remaining two octahedral sites are occupied by solvent molecules. It was found that the
complexed form only corresponds to one of the two diastereoisomers, 4'R, 2"R, 4"R-
pyochelin. This is also the case in another siderophore yersiniabactin 36, which forms
the Fe3+ complex only with one of its two stereoisomers containing the same R
configuration at those chiral centres as pyochelin.110,123
Enantio-pyochelin 76, the optical antipode of pyochelin produced from P. protegens, is
found to promote growth under iron-limiting conditions in P. protegens, thus acting as a
siderophore in the same way as pyochelin in P. aeruginosa. However, the interspecies
studies indicate that enantio-pyochelin did not promote growth in P. aeruginosa and
neither did pyochelin in P. protegens, suggesting that the pyochelin or enantio-
pyochelin-mediated iron uptake machinery is highly stereospecific in both species.114
Chapter 1 : Introduction
31
The reason underlying this is considered to prevent potential competitors occupying the
same ecological niche from stealing heterologous iron-siderophore complexes.124-125
Interestingly, pyochelin has been shown to have antibiotic activity against several
Xanthomonas species and S. aureus, which is not a result of iron competition, but due to
its ability to generate reactive oxygen species (ROS).126 However, analysis of the
spectrum of the antibacterial activity of pyochelin revealed that bacteria unable to
produce catecholate siderophores are sensitive to pyochelin while catecholate
siderophore producers are resistant by the ability of catecholate siderophores, such as
enterobactin 67 (Figure 1.16), to reduce ROS.126 In addition, pyochelin has been
investigated as a candidate for the Trojan horse prodrug strategy to develop potent
antibiotics by conjugation with fluoroquinolones.127
1.5.2 Biosynthesis of pyochelin and enantio-pyochelin
The biosynthesis of both pyochelin 74 in P. aeruginosa and enantio-pyochelin 76 in P.
protegens belongs to the NRPSs system, which has many similarities (Figure 1.19).
Figure 1.19. Organisation of the pyochelin biosynthetic gene clusters in P. protegens
CHA0, P. aeruginosa PAO1 and S. scabies 87-22. Pyochelin biosynthetic gene
homologues are colour-coded based on conserved function. Genes presumably involved
in pyochelin export are grey, while non-homologous genes are white. The pchE gene is
smaller in P. protegens CHA0 than in P. aeruginosa PAO1 due to the absence of an MT-
like coding region (coloured in black in pchE in P. aeruginosa PAO1).
Chapter 1 : Introduction
32
The biosynthetic genes for pyochelin 74 in P. aeruginosa are organised in two operons,
pchDCBA and pchEFGHI, with pchR encoding a transcriptional regulator. It has been
well established that pyochelin 74 is biosynthesized from salicylate and two molecules
of L-cysteine (Figure 1.19 and 1.20).128-133 Two enzymes, PchA, an isochorismate
synthase, and PchB, an isochorismate pyruvate-lyase, are responsible for the biosynthesis
of salicylate, which is then adenylated by the A domain, PchD. The gene pchC encodes
an external thioesterase, which belongs to the type II TE family.134 PchC has been
demonstrated in vivo to play a proofreading role by removing wrongly charged molecules
from the PCP domains of PchE and PchF.133 The adenylated salicylate is loaded onto the
N-terminal PCP domain of the peptide synthetase PchE and condensed with an activated
L-cysteine, which is linked at the C-terminus of PchE. This process, including
condensation, epimerisation and cyclisation reactions catalysed by PchE, leads to the
formation of the first thiazoline moiety. The peptide synthetase PchF catalyses
condensation of the PchE-bound 2-hydroxyphenylthiazolinyl intermediate with a second
molecule of L-cysteine. PchF catalyses an analogous reaction to PchE, resulting in the
second thiazoline ring. Reduction by the thiazoline reductase PchG and N-methylation
by the MT domain of PchF gives the pyochelin scaffold. It should be noted that the first
L-cysteine molecule incorporated is converted to its D-isoform by the methyl transferase
(MT)-like domain in PchE while the second cysteine retains its L-configuration, leading
to pyochelin 74 with 4'R and 4"R configuration on the thiazoline and thiazolidine rings,
respectively. The internal TE domain in PchF catalyses hydrolysis to release pyochelin
74 from the biosynthetic enzymes.
Chapter 1 : Introduction
33
Figure 1.20. Proposed pathway for the biosynthesis of pyochelin in P. aeruginosa.
The biosynthetic pathway of enantio-pyochelin 76 in P. protegens is very similar to
pyochelin 74 biosynthesis in P. aeruginosa, utilising homologues of the six proteins
PchABCDEF (Figure 1.19 and 1.21),114 although the organisation of the two biosynthetic
gene clusters is slightly different. Their main differences focus on the following two
aspects: i) PchK, which is proposed to be the thiazoline reductase responsible for
reduction of the second thiazoline, is not homologous to the reductase PchG that is
present in the P. aeruginosa pyochelin biosynthetic gene cluster. However, the
underlying mechanism of reduction (and epimerisation) by PchK, which leads to the
stereocenter with D-configuration in the second thiazoline ring, is still unknown, possibly
also involving the use of other enzyme(s). ii) The PchE homologue in P. protegens is
smaller than in P. aeruginosa due to the absence of an MT-like domain, which is believed
to be the epimerisation region responsible for creation of the D-configured stereocenter
in the first thiazoline of pyochelin. Collectively, these differences explain why P.
protegens produces enantio-pyochelin 76 with the opposite stereochemistry to pyochelin
74. It is noteworthy that the PchR regulators have been proposed to be one of the
Chapter 1 : Introduction
34
determinants for the stereospecificity of iron uptake mediated by pyochelin and enantio-
pyochelin in Pseudomonas species.135
Figure 1.21. Proposed pathway for the biosynthesis of enantio-pyochelin in P. protegens.
Pyochelin 74 is also produced by a Gram-positive bacterium, S. scabies 87-22, the
biosynthetic cluster of which contains six proteins, SCAB1381, SCAB1421, SCAB1411,
SCAB1481, SCAB1471 and SCAB1461, functionally related to PchABCDEFG in P.
aeruginosa.121 SCAB1381 is proposed to be the salicylate synthase responsible for the
formation of the salicylate moiety from chorismate. This is in contrast to the pathways
in Pseudomonas species, in which two enzymes PchA and PchB are required. The other
difference between two species is, a TetR-family protein (SCAB1401) and an AfsR-
family protein (SCAB1371) are involved in the transcriptional regulation of the
pyochelin biosynthetic gene cluster in S. scabies 87-22, whereas the Pseudomonas
species utilises an AraC-family protein (PchR).121
Chapter 1 : Introduction
35
1.5.3 Pulicatins, thiazostatin and watasemycin
Pulicatins A-E 77-81 (Figure 1.17) is a group of thiazoline metabolites isolated from a
cone snail Conus pulicarius associate, Streptomyces sp. CP32, and have been reported to
have neuroactive activities. The absolute stereochemistry of pulicatin A 77 and B 78 has
been fully elucidated based on studies using nuclear overhauser effect spectroscopy
(NOESY), circular dichroism (CD) and their derivatization to Mosher’s esters, namely
(R)-(+) or (S)-(-) α-methoxy-α-(trifluoromethyl) phenylacetyl ester.115 The configuration
of the stereogenic centre C-5' has been established as S as shown in Figure 1.17.
Thiazostatin 82 is a close analogue of pyochelin which differs only by an additional
methyl group at C-4". It was first reported in 1989 to be isolated from Streptomyces
tolurosus 1368-MT1 during screening for antioxidant substances.116 As other
2,4-disubstituted thiazolidine compounds, it presents as a pair of diastereoisomers due to
epimerisation at C-2", with antioxidant activity. However, thiazostatin was not reported
with antimicrobial or iron-chelating activity since its discovery, probably due to lack of
published data on this compound.
Watasemycin 83, which possesses a methyl group at C-5' of its thiazoline ring instead of
a hydrogen atom, is structurally closely related to thiazostatin and pyochelin with similar
epimerisation at C-2". This compound was first reported in 2002 by Furumai et al. to be
isolated from Streptomyces sp. TP-A0597, accompanied by isolation of thiazostatin
82.117 Watasemycin has been reported with modest antibiotic activity against Gram-
positive and negative bacteria and yeast. Recently, watasemycin and thiazostatin, as well
as some other 2-hydroxylphenylthiazolines, have been identified to be the metabolic
products of a biosynthetic gene cluster from S. venezuelae ATCC10712.136 More details
about investigations into watasemycin biosynthesis in S. venezuelae are described in
Chapter 4.
The relative stereochemistry of thiazostatin and watasemycin has been proposed based
on their NOESY studies. The proposed configuration for thiazostatin is 4'S, 4"R or 4'R,
4"S while for watasemycin is 4'S, 5'S, 4"R or 4'R, 5'R, 4"S (Figure 1.22).117 However,
their absolute configuration hasn’t been determined yet.
Chapter 1 : Introduction
36
Figure 1.22. Reported relative stereochemistry of watasemycin and thiazostatin.
1.6 S-Adenosylmethionine (SAM)-dependent methylation
S-Adenosylmethionine (SAM) is a versatile small molecule used in many biological
reactions (Figure 1.23). It is a commonly used methyl donor in numerous biologically
important methylation reactions, including DNA, RNA and protein methylation, as well
as in the biosynthesis of bacterial specialised metabolites. The 3-amino-3-carboxypropyl
group of SAM can also be transferred to different acceptor molecules, such as RNA and
proteins. In addition to serving as a methyl donor and a 3-amino-3-carboxypropyl donor,
SAM is also involved in several other types of group transfer reactions, e.g., the adenosyl
group transfer in fluorinase and chlorinase.137-139
Figure 1.23. Three typical SAM-dependent reactions with methyl, 3-amino-3-
carboxypropyl or adenosyl group transfer.
In classical SAM-dependent methylation reactions, a simple SN2-type nucleophilic
substitution mechanism is utilised, in which a nucleophile such as a heteroatom (N, O or
Chapter 1 : Introduction
37
S), directly attacks the methyl group attached to the positively charged sulfonium ion in
SAM, with concomitant formation of S-adenosyl homocysteine (SAH). However, SAM-
dependent C-methylations are more complicated and have been proposed to involve
different mechanisms. Substrates that are more nucleophilic or easily converted to a
nucleophile can be methylated via simple nucleophilic mechanisms, for example, the
SAM-dependent formation of cyclopropane on unsaturated fatty acids in bacteria or
enzyme-assisted nucleophile formation in DNA cytosine C-5 methylation.140 The methyl
group at C-4" of watasemycin and thiazostatin is presumably generated by nucleophilic
methylation of the α-carbon of cysteine with SAM.117 The mechanism of introducing a
methyl group at the inert C-5' position of watasemycin is unclear due to the fact that
nucleophilic substitution is not likely to take place at such an unreactive carbon centre.
More recently, it has become clear that SAM is also able to methylate inert carbon or
phosphorous atoms via radical intermediates, which probably involves the use of
additional cofactors. These reactions are known to be catalysed by radical SAM (RS)
enzymes,141 which initiate a diverse set of reactions via generation of a radical
intermediate. One molecule of SAM receives an electron transferred from the reduced
[4Fe-4S] cluster and undergoes homolytic cleavage of the C5'-S bond to generate the
5′-deoxyadenosyl (5′-dAdo) radical (Figure 1.24). This initiation step is common to all
RS methylases. Several reviews have summarised these RS methylases and classified
them into different groups, mainly depending on their protein architectures, cofactor
requirements and mechanisms of catalysis.141-143 So far, five classes of RS methylases
have been described, including class A, B, C, D and a new class E, which was identified
very recently.142-144
Figure 1.24. Generation of 5′-dAdo radical intermediate from SAM as an initiation step
for methylation reactions catalysed by RS methylases.
Class A RS methylases are characterised by two strictly conserved cysteine residues and
methylation is carried out by a protein-based methylene radical, resulting in only two of
Chapter 1 : Introduction
38
the three protons from the methyl group of SAM being transferred to the product. These
methyltransferases perform methylations on sp2-hybridised carbon centres and are
commonly involved in RNA methylation, e.g., RlmN and Cfr,145-146 which are best
characterised examples of this class to catalyse methylation on adenosine of RNAs.
Class C methylases are annotated with a second domain having high sequence similarity
to coproporphyrinogen III oxidase, in addition to the SAM-binding domain, and catalyse
methylation of sp2-hybridised carbon atoms in aromatic heterocycles. These enzymes
appear to be involved in the biosynthesis of many antitumor and antibiotic bacterial
metabolites such as bleomycin and nosiheptide.141
The only example of class D enzyme is MJ0619 in methanopterin biosynthesis, which
uses methylenetetrahydrofolate as an alternative methyl group source.147 A family of
NifB proteins have been reported by Hu et al. to catalyse the essential step of radical
SAM-dependent carbide insertion in the assembly of nitrogenase metallocofactor.144
This group of enzymes has been treated as a new class of RS enzymes, the class E
methylases.
1.6.1 Class B radical SAM methylases
Class B RS methylases are of particular interest among all RS methylases due to their
ability to catalyse methylation on phosphinate phosphorous atoms and sp2- and sp3-
hybridised carbon centres, representing the largest and most versatile class of RS
methylases.141,148 The methylation reactions catalysed by class B enzymes are involved
in the biosynthesis of a wide variety of microbial natural products, including
aminoglycosides (e.g., gentamycin G418 85 and fortimicin A 86),149-150 β-lactams (e.g.,
thienamycin 87),151 polyketides (e.g., chondrochloren B 88),152 ribosomally
biosynthesised and posttranslationally modified peptides (RiPPs) (e.g., thiostrepton A 89,
bottromycin A2 90 and polytheonamide 91),153-158 nonribosomally biosynthesised
peptides (NRPs) (e.g., watasemycin 83),136 phosphonates (e.g., phosphinothricin 10),159
enediynes (e.g., calicheamicin 92 and kedarcidin 93),160-161 aminocoumarins (e.g.,
clorobiocin 94) and terpenes (e.g., 2-methylhopene 95) (Figure 1.25).162-163
Chapter 1 : Introduction
39
Figure 1.25. Specialised metabolites for which class B RS methylases are known or
predicted to be involved in their biosynthesis, with corresponding functional groups
coloured in red.
Chapter 1 : Introduction
40
These enzymes are distinguished in architecture from other classes of RS methylases by
an N-terminal methylcobalamin (MeCbl)-binding domain in addition to the core radical
SAM domain. The most common reactions catalysed by class B RS methylases involve
the methylation of un-activated sp3-hybridised carbon centres. Examples include the
elongation of a methoxy group to an ethoxy group, catalysed by CndI in
chondrochloren B 88 biosynthesis, and the β-position methylation of the valine moiety
catalysed by a putative class B enzyme in pheganomycin 96 biosynthesis (Figure
1.25).152,164
The catalytic mechanism for class B RS methylases is proposed as shown in Figure 1.26.
Same as other RS methylases, the 5'-dAdo radical is firstly generated from the common
initiation step utilising a [4Fe-4S] cluster. Then the 5'-dAdo radical abstracts a hydrogen
atom from the target carbon centre to give a substrate radical, which attacks the methyl
moiety of MeCbl, leading to homolytic cleavage of the C-Co bond. This results in the
production of cob(II)alamin and the product modified at the methylating position. The
inactive cob(II)alamin can be returned to MeCbl by obtaining an electron and
subsequently reacting with another molecule of SAM.
Figure 1.26. Proposed catalytic mechanism of class B RS methylases (left) and the
structure of MeCbl (right).143,148
1.6.1.1 Stereochemical course of C-methylation by class B radical SAM methylases
It has been proposed that C-methylations catalysed by class B RS methylases proceed
with inversion of configuration at the reacting carbon centre.148 Clear examples of this
process include the GenD1-catalysed methylation at C-4" in gentamicin biosynthesis and
Chapter 1 : Introduction
41
the methylation by MoeK5 in moenomycin 97 biosynthesis (Figure 1.25).149,165 The
inverted stereochemistry suggests that SAM and MeCbl are positioned on opposite sides
of the substrate in the active site of class B RS methylases (Figure 1.27).148 Such an
arrangement dictates that the substrate radical must undergo inversion of configuration
before it can attack MeCbl. Further details using stereospecifically isotope-labelled
precursors to decipher the stereochemical course are discussed in Chapter 4.
Figure 1.27. Proposed organisation of the active site in class B RS methylases.148
Chapter 1 : Introduction
42
1.7 Aims of project
This work had three main aims: i) to establish the role of MmfL in AHFCA assembly in
S. coelicolor A3(2); ii) to study the regulation mechanism of specialised metabolite
biosynthesis by AHFCAs and ArpA-like repressor proteins in Streptomyces; iii) to
characterise the watasemycin biosynthetic pathway in S. venezuelae ATCC10712. The
following objectives were pursued.
Investigate the function of MmfL in vitro for AHFCA assembly by synthesising
a synthetic standard of the putative dephosphorylated butenolide product of the
MmfL-catalysed reaction.
Study the in vitro interaction between MmfR, its cognate operator sequences and
AHFCAs or SCB1 using EMSAs.
Identify the key residues involved in the ligand recognition by MmfR using
site-directed mutagenesis based on analysis of the X-ray crystal structure of
MmfR with AHFCA2.
Investigate the in vitro interaction between the MmfR homologue SgnR, its
cognate operator sequences and corresponding AHFCA signalling molecules
from the gaburedin biosynthetic gene cluster in S. venezuelae.
Identification and characterisation of the pyochelin-like metabolic product from
the watasemycin biosynthetic gene cluster in S. venezuelae ATCC10712.
Establish the function of several enzymes, including Sven0508, Sven0516 and
Sven0515, in the biosynthesis of 2-hydroxylphenylthiazolines in S. venezuelae.
Study the stereochemical course of the methylation at C-5' in watasemycin
biosynthesis.
Chapter 2 : MmfL in AHFCA biosynthesis
43
2. Investigation of the role of MmfL in AHFCA biosynthesis
Chapter 2 : MmfL in AHFCA biosynthesis
44
2.1 Introduction to previous studies
MmfL, MmfP and MmfH are a cassette of three enzymes proposed to be responsible for
the biosynthesis of AHFCA signalling molecules that control expression of the
methylenomycin biosynthetic gene clsuster, as discussed in Section 1.2.3. During the
biosynthetic process, MmfL is proposed to catalyse an analogous reaction to AfsA,67
including the condensation of various ACP-bound β-ketothioester intermediates in fatty
acid biosynthesis with DHAP derived from glycolysis, followed by spontaneous
intramolecular aldol condensation and dehydration, resulting in the corresponding
phosphorylated butenolides. Subsequent dephosphorylation of the butenolides catalysed
by MmfP, followed by MmfH-catalysed oxidation of the butenolide would yield an
acyclic aldehyde intermediate, which undergoes cyclisation, dehydration and reduction
to yield the AHFCAs (Scheme 2.1).
Scheme 2.1. Proposed biosynthetic pathway for AHFCAs, involving different
β-ketoacyl-ACP thioesters.
Based on the sequence similarity between MmfL and AfsA as shown in Figure 1.8 in
Chapter 1, as well as the work reported by Kato et al.,67 the role of MmfL in AHFCA
biosynthesis may be similar to AfsA in the γ-butyrolactone A-factor biosynthesis in S.
griseus. During investigation into the function of AfsA, Kato and co-workers used an 8-
methyl-3-oxononanoyl-NAC 98 to mimic the corresponding β-ketoacyl-ACP as the
substrate for AfsA in vitro (Scheme 2.2). The enzymatic reaction of AfsA with NAC β-
ketothioester 98 and DHAP 50 results in the 8-methyl-3-oxononanoyl-DHAP ester 52,
which was confirmed by comparison to a synthetic standard. The product of the AfsA
reaction is proposed to undergo an aldol condensation and dehydration nonenzymatically
Chapter 2 : MmfL in AHFCA biosynthesis
45
to afford the phosphorylated butenolide 55, which can be observed by LC-MS as the
dephosphorylated form 54, after phosphatase treatment.
Scheme 2.2. Enzyme assays performed by Kato and co-workers for investigation of
AfsA function.67
Initial investigations into the biosynthetic pathway of AHFCAs were undertaken in order
to establish the specific roles of MmfL, MmfP and MmfH in AHFCA assembly (work
by N. Malet).77 The three enzymes were overproduced in E. coli and purified as
recombinant N-terminal His6-tagged proteins. The reaction catalysed by MmfL is
proposed to be the first step in AHFCA biosynthesis, resulting in a phosphorylated
butenolide intermediate.
Enzymatic reactions were performed for His6-MmfL using a synthetic substrate, the
NAC β-ketothioester 61, which is a mimic of the proposed β-ketoacyl-ACP thioester
precursor of AHFCA1 (Scheme 2.3). The MmfP-mediated dephosphorylation could be
replaced by treatment with alkaline phosphatase in vitro, similar to that used in the AfsA
study. The putative phosphorylated butenolide product 59 could not be detected due to
its high polarity. Further treatment of 59 with alkaline phosphatase yields a
dephosphorylated product, which can be observed by LC-MS. However, structure
elucidation of this product failed in previous work. On one hand, purification of the
putative butenolide 60 for subsequent NMR spectroscopic analysis from a scaled-up
enzymatic reaction appears extremely difficult because of its inherent instability. On the
other hand, attempts to synthesise the butenolide 60 as a reference also failed.
Chapter 2 : MmfL in AHFCA biosynthesis
46
Scheme 2.3. Enzymatic reactions performed by N. Malet with synthetic 5-methyl-3-
oxohexanoyl-NAC for investigation of the function of MmfL.77
In order to determine the specific biosynthetic role of MmfL in the assembly of AHFCAs,
further studies need to be carried out, mainly focusing on structure elucidation of the
presumed dephosphorylated butenolide. The key point was preparation of an authentic
standard of butenolide to use it as a reference for characterisation of the product from an
MmfL-catalysed reaction.
2.2 Synthesis of the N-acetylcysteamine (NAC) β-ketothioester substrate analogue
The substrates of MmfL are predicted to be DHAP, which is a commercially available
compound, and various ACP-bound β-ketothioesters. Two NAC β-ketothioesters with
different alkyl chains 61 and 101, corresponding to the β-ketoacyl-ACP thioester
precursors of AHFCA1 and AHFCA5, respectively, were prepared in this study as
substrate analogues for MmfL (Scheme 2.4).
Scheme 2.4. Synthesis of NAC β-ketothioesters 61 and 101 as substrate analogues of
MmfL.
Synthetic procedures for such NAC β-ketothioesters are well described in the
literature.67,77,166 They typically involve the acylation of Meldrum’s acid followed by a
reaction with NAC to yield the desired β-ketothioester. Due to availability of the acid or
acid chloride in our lab, 3-methylbutyryl chloride and hexanoic acid were used for
Chapter 2 : MmfL in AHFCA biosynthesis
47
synthesis of the corresponding acyl Meldrum’s acids, which led to the NAC
β-ketothioester precursors of AHFCA1 and AHFCA5, respectively. It should be noted
that the prepared NAC β-ketothioesters exist in a mixture of the keto and enol form due
to their tendency for keto-enol tautomerism, which can be easily distinguished from
1H NMR spectra.
The combination of 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC) and 4-
dimethylaminopyridine (DMAP) was used to couple the carboxylic acid to Meldrum’s
acid,167 while triethylamine and DMAP were adopted for acylation with acid chloride.
EDC is capable of activating the carboxylic acid groups and offers facile removal of the
urea by-product by aqueous washing due to the pendant dimethylamine moiety
bestowing water-solubility. Moreover, coupling of the carboxylic acid with Meldrum’s
acid by EDC simplifies the preparation of acyl Meldrum’s acid since the acylation takes
one step, comparatively, acid chloride formation and subsequent acylation requires two
steps. The general mechanism for the coupling of the carboxylic acid with Meldrum’s
acid induced by EDC and DMAP is described in Scheme 2.5.
Scheme 2.5. General mechanism for the coupling of the carboxylic acid with Meldrum’s
acid using EDC and DMAP.
2.3 Synthesis of the butenolide authentic standard
In order to confirm the proposed butenolide structure of the dephosphorylated product,
the synthesis of an authentic standard of 60 was required. The tert-butyldimethylsilyl
Chapter 2 : MmfL in AHFCA biosynthesis
48
(TBDMS) protected butenolide 105 was chemically synthesised by N. Malet but he was
unable to complete the final deprotection step as shown in Scheme 2.6.
Scheme 2.6. Attempted synthesis of the dephosphorylated putative product 60 of the
MmfL-catalysed condensation of DHAP and NAC β-ketothioester 61 by N. Malet.77
Additionally, although the synthetic approach was able to yield the TBDMS-protected
butenolide 105, there is still room for improvement in the synthesis. Firstly, hydrolysis
of the β-ketoester to form the β-ketoacid 103 had a low yield with a long reaction time
of nearly five days, and considerable amount of starting β-ketoester remained. Secondly,
for commercially unavailable β-ketoesters, additional synthetic steps would be required
to prepare them.
Scheme 2.7. Synthesis of the TBDMS-protected butenolide 106.
In order to improve the synthesis of the TBDMS-protected butenolide, an alternative
synthetic route was adopted in this study as shown in Scheme 2.7. This synthetic
methodology was reported by Sello and co-workers for the synthesis of A-factor,168
Chapter 2 : MmfL in AHFCA biosynthesis
49
which was also used in the synthesis of γ-butyrolactone SCB1 in this work as described
in Section 3.5.1.
The synthesis proceeds via a TBDMS-protected butenolide intermediate. The
appropriate acyl Meldrum’s acid is refluxed in toluene in the presence of TBDMS mono-
protected dihydroxyacetone, yielding the corresponding β-ketoester, which cyclises
during purification by flash chromatography on silica. This key step involves a tandem
reaction sequence of an esterification and intramolecular Knoevenagel condensation.
During the synthesis, it was noticed that the TBDMS-protected butenolides are not very
stable and may deteriorate during storage even at -20 °C, particularly for those with
shorter and branched alkyl chains such as the protected butenolide 105 with an isopropyl
side chain corresponding to AHFCA1. This also explains the reason for the difficult
purification during preparation of these compounds. Thus most of our efforts were
focused on the TBDMS-protected butenolide 106 corresponding to AHFCA5, which is
an unbranched pentyl group on the side chain.
As the instability of the butenolide structure is well known from reported literature and
previous studies,77,169 attempts were focused on deprotection of the silyl group using very
mild conditions. Initial studies tested the effectiveness of tetrabutylammonium fluoride
(TBAF) as a desilylating agent to remove the protecting group (Scheme 2.6). The
reaction with TBAF was performed in THF at 0 °C for 30 min (Scheme 2.8). LC-MS
was used to analyse reaction products because of the rapid degradation of the butenolide
product. Only a tiny amount of the peak with m/z corresponding to the deprotected
butenolide 107 was detected by LC-MS. This is probably due to generation of strongly
basic ammonium alkoxides during TBAF treatment that are incompatible with base
sensitive compounds.
We then explored the feasibility of using hydrogen fluoride-pyridine complex, which is
a milder reaction condition (Scheme 2.8). The deprotection was performed in THF at
0 °C for 40 min. It resulted in a much higher peak at m/z = 235.0941 on LC-MS,
corresponding to [M+Na]+ of the deprotected butenolide 107, as shown in trace A of
Figure 2.1.
Chapter 2 : MmfL in AHFCA biosynthesis
50
Scheme 2.8. Attempts to deprotect the TBDMS-protected butenolide 106.
The instability of protected butenolide 106 is clearly demonstrated by the fact that a
sample of it showed a low level of deprotected butenolide 107 by LC-MS (trace B of
Figure 2.1). This highlights the importance of conducting LC-MS analysis with fresh
samples in the studies related to the catalytic activity of MmfL.
Figure 2.1. EICs at m/z = 235.0941 (A and B), corresponding to [M+Na]+ for butenolide
107, (A) from LC-MS analysis of the deprotection by hydrogen fluoride-pyridine, (B)
from a sample of the synthetic TBDMS-protected butenolide 106, and m/z = 349.1806
(C), corresponding to [M+Na]+ for TBDMS-protected butenolide 106.
Since NMR characterisation of 107 was not possible, MS/MS was used to further
characterise the compound. The fragmentation pattern resulting from the [M+H]+ parent
ion in the ESI-MS/MS spectrum of the synthetic butenolide 107 from deprotection by
hydrogen fluoride-pyridine complex correlates well with its predicted fragmentation as
shown in Figure 2.2. Thus we conclude that the deprotection reaction from the TBDMS-
protected butenolide 106 by hydrogen fluoride-pyridine produces butenolide 107.
Chapter 2 : MmfL in AHFCA biosynthesis
51
Figure 2.2. Positive ion mode ESI-MS/MS spectrum of the synthetic butenolide 107
from deprotection by hydrogen fluoride-pyridine complex with predicted fragmentation
annotated.
2.4 Overproduction, purification and characterisation of recombinant MmfL
A glycerol stock of E. coli BL21 star (DE3) cells containing the pET151-mmfL plasmid
was obtained from Nicolas Malet. His6-MmfL was purified by nickel affinity
chromatography as described in Section 6.3.16. SDS-PAGE analysis of the purified His6-
MmfL protein (~41 kDa) confirmed it had the expected molecular weight (Figure 2.3).
Figure 2.3. SDS-PAGE analysis of His6-MmfL (~41 kDa) after purification. (MWM:
molecular weight marker)
The molecular weight of His6- MmfL was further confirmed by LC-HRMS (Figure 2.4).
Chapter 2 : MmfL in AHFCA biosynthesis
52
Figure 2.4. Measured (top) and deconvoluted (bottom) mass spectra of His6-MmfL
(calculated mass = 41442.0 Da).
2.5 Characterisation of dephosphorylated product of MmfL-catalysed reaction
Based on previous studies, MmfL-catalysed condensation of DHAP and the NAC β-
ketothioester 101, which is a mimic of the proposed β-ketoacyl-ACP thioester precursor
of AHFCA5, was followed by treatment with alkaline phosphatase (Scheme 2.9). In this
study, it was also found that a good temperature for MmfL-catalysed reaction and
subsequent dephosphorylation is 25 °C, instead of 30 °C used by Nicolas Malet.
Scheme 2.9. MmfL-catalysed condensation of DHAP and NAC β-ketothioester 101,
followed by dephosphorylation to form butenolide 107.
Reactions were quenched with methanol and the supernatants were analysed by LC-
HRMS. LC-HRMS analyses in positive ion mode of the condensation and subsequent
dephosphorylation reactions revealed a product that gave rise to an ion with m/z 235.0936,
corresponding to [M+Na]+ of butenolide 107. This product was absent from negative
controls, confirming that it results from MmfL-catalysed condensation of DHAP and
NAC β-ketothioester 101 (Figure 2.5).
Chapter 2 : MmfL in AHFCA biosynthesis
53
Figure 2.5. EICs for m/z = 235.0941 (corresponding to [M+Na]+ for butenolide 107)
from LC-MS analyses of the dephosphorylated product of the MmfL-catalysed reaction
of DHAP with NAC β-ketothioester 101.
The molecular formula of the dephosphorylated product of the MmfL-catalysed
condensation of DHAP and NAC β-ketothioester 101 are identical to the molecular
formula of AHFCA5 44. In order to distinguish between the two species, LC-MS-guided
comparisons of the dephosphorylated product of the MmfL-catalysed condensation with
synthetic AHFCA5, which was obtained from Nicolas Malet, were carried out. The
results from LC-MS shown in Figure 2.6 confirmed that they are different.
Chapter 2 : MmfL in AHFCA biosynthesis
54
Figure 2.6. EICs for m/z = 235.0941 from LC-MS analyses of the dephosphorylated
product of the MmfL-catalysed reaction of DHAP with NAC β-ketothioester 101 and
synthetic AHFCA5 44, as well as a co-injection of them.
Figure 2.7. (A) EICs for m/z = 235.0941 from LC-MS analyses of the dephosphorylated
product of the MmfL-catalysed reaction of DHAP with NAC β-ketothioester 101 and
synthetic butenolide 107, as well as a co-injection of them in equal volumes.
(B) Comparison of the MS/MS fragmentation spectra of the peaks with m/z = 213
(corresponding to [M+H]+) obtained from synthetic butenolide 107 and the enzymatic
reaction product. Precursor ions are indicated with diamonds.
Chapter 2 : MmfL in AHFCA biosynthesis
55
Further LC-MS comparisons of the dephosphorylated product of the MmfL-catalysed
condensation of DHAP and NAC β-ketothioester 101 were carried out with synthetic
butenolide 107 as an authentic standard. The product from enzymatic reactions and
synthetic butenolide 107 had the same retention time and the same MS/MS fragmentation
pattern as shown in Figure 2.7. This strongly suggests that the dephosphorylated product
of the MmfL-catalysed reaction is actually the presumed butenolide.
The condensation/dephosphorylation reaction cascade was modestly scaled up in an
attempt to chemically protect the hydroxyl group of the butenolide product 107 with a
silyl group. The enzymatic product was extracted with dichloromethane and dried over
MgSO4 before trying the silyl protection reaction with TBDMSCl and imidazole.
However, the TBDMS-protected butenolide 106 was not detected from LC-MS analysis
of the sample. Attempts with a very strong silylating agent, tert-butyldimethylsilyl
triflate (TBDMSOTf), and 2,6-lutidine also failed, presumably due to the instability of
the butenolide 107 under basic conditions of the silylating reactions (Scheme 2.10).
Scheme 2.10. Attempted synthesis of the TBDMS-protected butenolide 106 from the
dephosphorylated product of the MmfL-catalysed condensation of DHAP and NAC β-
ketothioester 101.
Chapter 2 : MmfL in AHFCA biosynthesis
56
2.6 Summary
MmfL was proposed to catalyse an analogous reaction to AfsA, involving condensation
of ACP-bound β-ketothioester with DHAP followed by spontaneous intramolecular aldol
condensation and dehydration to form a butenolide intermediate. Previous in vitro studies
with recombinant His6-MmfL protein failed to provide enough evidence for structure
elucidation of the butenolide.
The synthesis of a butenolide 107 with its side chain corresponding to that of AHFCA5
was achieved in this study. It involves a tandem reaction sequence of an esterification
and intramolecular Knoevenagel condensation from acyl Meldrum’s acid 100 and
TBDMS mono-protected dihydroxyacetone 102. After preparation of the TBDMS-
protected butenolide 106, a mild deprotection condition using hydrogen fluoride-
pyridine complex was adopted to give the desired butenolide 107, as evidenced by its
ESI-MS/MS spectrum. Moreover, LC-MS analysis provided evidence that the TBDMS-
protected butenolide compounds are not very stable, leading to a mixture with its
deprotected butenolide form.
The dephosphorylated product of the MmfL-catalysed reaction was confirmed to be
butenolide from LC-MS/MS comparison with synthetic butenolide 107 as an authentic
standard. However, TBDMS protection of the dephosphorylated product of the MmfL-
catalysed reaction proved to be difficult, possibly due to the instability of the butenolide
structure under basic conditions required for silylating reactions.
Based on current studies, it has been demonstrated that MmfL catalyses the reaction
between ACP-bound β-ketothioester and DHAP to form a phosphorylated butenolide
intermediate. This process is analogous to that catalysed by AfsA during the
γ-butyrolactone signalling molecule A-factor biosynthesis in S. griseus.
Chapter 3 : Interactions between DNA, ligand and protein
57
3. Interactions between DNA, ligand and ArpA-like
repressors
Chapter 3 : Interactions between DNA, ligand and protein
58
3.1 Introduction
A widely used method for studying DNA-protein interactions is electrophoretic mobility
shift assays (EMSAs), which are also known as gel shift assays or gel retardation
assays.170-171 Because free DNA and DNA-protein complexes migrate differently during
gel electrophoresis, they can be separated and detected on polyacrylamide or agarose gels.
Initial EMSAs were carried out for studying the interactions between MmfR and MmyR,
DNA sequences and different AHFCA compounds (work by P. Harrison). Although it is
common to use labelled DNA that can be visualised via interaction with a corresponding
antibody, this approach led to high levels of background. It was found that simply
staining with GelRed and visualisation with UV light gave a clearer gel, although more
DNA was required due to lower sensitivity of this method.
Some details for running EMSAs were concluded from previous work. Firstly, the non-
denaturing polyacrylamide gels are better made fresh and pre-run before loading samples.
Pre-running the gels distributes ions equally within the gel and removes un-polymerised
acrylamide, salts and ammonium persulfate from the gel, which can interfere with the
progression of the sample through the gel; secondly, it is better to prepare the protein in
gel filtration buffer which does not contain imidazole. Imidazole salts contained in the
protein buffer can cause streaking of the DNA through the gel, possibly due to formation
of salt-DNA complexes; finally, ligands should be dissolved in DMSO as AHFCAs
solubilised in methanol produced inconsistent results.
A schematic representation of the EMSAs used in this study is shown in Figure 3.1.
DNA-protein binding can be identified by running parallel samples with and without
protein. If the protein binds to the DNA, there will be retardation of the DNA migration
through the gel, resulting in a band shift. Upon addition of an active ligand, the receptor
protein interacts with the ligand, leading to a conformational change. This results in a
significant decrease of the DNA-binding affinity for the protein. Therefore, the DNA will
be released from the protein and its migration through the gel will no longer be retarded.
Chapter 3 : Interactions between DNA, ligand and protein
59
Figure 3.1. Schematic representation of EMSAs in this study.
EMSAs previously performed for MmfR/MmyR included two DNA fragments, one of
which is the entire mmfR-mmfL intergenic region (194 bp) containing MARE1 and the
other is a shorter mmfR-mmfL intergenic sequence (44 bp) with an additional 13 bp
nucleotides flanking either side. An example of the previous EMSAs is shown in
Figure 3.2. It has been confirmed that MmfR protein can bind to both the DNA fragment
of the entire mmfR-mmfL intergenic region and the shorter 44-bp sequence. This indicates
that the MARE1 sequence located between mmfR and mmfL is one of the binding sites
for MmfR. However, due to solubility issues with MmyR, its binding to either of these
two DNA fragments was not observed in EMSAs.
Figure 3.2. An example of the EMSAs with DNA fragment of the entire mmfR-mmfL
intergenic region (194 bp), MmfR and AHFCA1 by P. Harrison (left), and the S.
coelicolor methylenomycin biosynthetic gene cluster with location of three proposed
MAREs highlighted in magenta (right).
Additionally, EMSAs were used to probe the ligand specificity of MmfR for different
AHFCA compounds. These in vitro data, combined with the results from the in vivo
experiments for induction of methylenomycin production (work by N. Malet), revealed
some important structural features for AHFCAs to interact with MmfR and cause its
Chapter 3 : Interactions between DNA, ligand and protein
60
release from the DNA.77 It has been showed that AHFCAs with alkyl chains between 3
and 5 carbons long are most effective at attenuating the ability of MmfR to bind the DNA
fragment.
Preliminary steps have been taken to understand the function of the transcriptional
regulator MmfR in regulation of methylenomycin biosynthesis in S. coelicolor A3(2).
However, further study of the interactions between MmfR, the MARE sequences and
signalling molecules is required. In order to study the DNA binding of MmfR more
specifically, Section 3.2 will focus on interactions between MmfR and a smaller DNA
fragment containing MARE1, as well as shorter oligonucleotides containing the other
two MARE sequences. Section 3.3 will introduce a comparison of the interactions of
MmfR with the five natural AHFCAs 1-5 in the presence of the DNA sequence
containing MARE1. As mentioned in Section 1.3.3, the X-ray crystal structure of MmfR
with AHFCA2 bound has been solved by the Corre and Challis groups. The EMSA
results with various MmfR mutants based on site-directed mutagenesis will be described
in Section 3.4. Section 3.5 will focus on the interaction of DNA and MmfR in the
presence of the synthetic SCB1 compared with the presence of AHFCA1. In addition,
the interactions between SgnR, a close homologue of MmfR in S. venezuelae, and its
proposed ARE sequences, as well as their interaction with a cognate ligand, will be
summarised in Section 3.6.
3.2 Interactions of MmfR with three MARE sequences
3.2.1. Confirmation of MmfR binding to MARE sequences
The pET151/D-TOPO plasmid containing mmfR was provided by Dr. Christophe Corre.
Chemically-competent E. coli BL21 star (DE3) cells were transformed with the pET151-
mmfR plasmid and the recombinant His6-MmfR protein was overproduced and purified
by nickel affinity chromatography as described in Section 6.3.16. SDS-PAGE analysis
of the purified His6-MmfR protein (~27.8 kDa) confirmed it had the expected molecular
weight (Figure 3.3).
Chapter 3 : Interactions between DNA, ligand and protein
61
Figure 3.3. SDS-PAGE analysis of His6-MmfR (~27.8 kDa) after purification. (MWM:
molecular weight marker)
The molecular weight of His6-MmfR was further confirmed by LC-HRMS (Figure 3.4).
Figure 3.4. Measured (top) and deconvoluted (bottom) mass spectra of His6-MmfR
(calculated mass = 27835.5 Da).
As discussed, the DNA fragments used previously are both considerably longer than the
18 bp MARE sequences shown in Table 1.1. Thus the EMSAs were repeated with shorter
DNA sequences (both single- or double-stranded). The nearly palindromic character of
the MARE sequences intrigued us. To check whether the palindrome is required for
MmfR recognition and binding, we first used a single-stranded DNA fragment, the 24 bp
perfectly palindromic sequence, 5'-AACATACCTTCGCGAAGGTATGTT-3', in the
EMSA with MmfR. Attempts with 6% non-denaturing polyacrylamide gels and the same
amount of DNA (0.1 pmol) previously used all failed. This is likely due to improper pore
size of the 6% gel, which is possibly too big for such small single-stranded
oligonucleotides. Also, the short DNA may be lower than the detection limit of the DNA
Chapter 3 : Interactions between DNA, ligand and protein
62
staining and visualisation method used in our EMSAs. The free DNA band could only
be observed after the acrylamide concentration of the gel was increased to 20% with the
use of 10 times more DNA fragment (1 pmol). However, no band shift was observed
with this single-stranded oligonucleotide or the annealed double-stranded palindrome
when MmfR was added, even after addition of 100 times more than DNA.
In these attempts, two slightly longer palindromic sequences,
5’-CAACATACCTTCGCGAAGGTATGTTG-3’ (26 bp) and
5’-GCAACATACCTTCGCGAAGGTATGTTGC-3’ (28 bp), also failed to show
interaction with MmfR in both the single-stranded and annealed form. It seems that this
perfectly palindromic sequence cannot be bound by MmfR. However, since the double-
stranded palindromic sequence was annealed from the single-stranded oligonucleotides
without purification, the conclusion needs to be taken with caution.
To improve the EMSAs with small DNA fragments, we then elected to use self-annealing
DNA fragments that can form hairpin sequences (Figure 3.5). The 24 bp oligonucleotide
in the middle of the hairpin structure is considerably flexible,172 allowing us to locate
the three 18-bp MARE sequences with an additional 3 nucleotides flanking either side.
The variable region is surrounded by GC clamps on both the 5' and 3' ends. The
sequences were ordered as single-stranded oligonucleotides that annealed to form the
hairpin structure (see Section 6.3.18).
Figure 3.5. Hairpin DNA containing three MAREs (18 bp, coloured in red) for EMSAs.
In view of the still small size of these DNA fragments, attempts were first carried out to
optimise the amount of DNA and MmfR used in each reaction and the acrylamide
concentration of the gel in order to obtain EMSA images with clear band shifts. A higher
amount of DNA fragments, 0.8 pmol, was determined to be appropriate in conjunction
with accordingly increased amount of protein, 4.0 pmol. Meanwhile, 10% non-
Chapter 3 : Interactions between DNA, ligand and protein
63
denaturing polyacrylamide gels showed the best images, although the time of gel
electrophoresis had to be increased to more than two hours.
These modifications meant the EMSAs using the hairpin DNA fragment containing the
MARE1 sequence worked much better than previously, as shown in Figure 3.6.
Moreover, DNA release from MmfR after addition of its natural ligand AHFCA1 was
successfully observed. The release profile of DNA from protein was consistent with that
observed with the 194-bp DNA of the entire mmfR-mmfL intergenic region.
Figure 3.6. Interaction of MmfR with a self-annealing DNA fragment forming a hairpin
sequence containing MARE1 located in the mmfL-mmfR intergenic region in response to
increasing amounts of AHFCA1. Lane 1: DNA fragment only. Lanes 2 to 7: AHFCA1
at 0, 0.8, 8, 20, 40 and 100 nmol respectively. Amounts of DNA and proteins were kept
constant (0.8 pmol and 4.0 pmol respectively).
The same conditions were then used for EMSAs with hairpin DNA containing the other
two MARE sequences, as shown in Figure 3.7. Though the progression of samples was
distorted, leading to skewed images, there are obvious band shifts due to formation of
the DNA-MmfR complex. This strongly indicated the ability of MmfR to bind the other
two specific MARE sequences. In the presence of the ligand AHFCA1, increasing
release of DNA fragments can also be observed. Thus, the three proposed MARE
sequences have been experimentally confirmed to be the binding sites for MmfR and
could be released upon the interaction of MmfR with AHFCA ligands.
Chapter 3 : Interactions between DNA, ligand and protein
64
Figure 3.7. Interaction of MmfR with self-annealing DNA fragments forming hairpin
sequences containing MARE2 located in the mmyY-mmyB intergenic region (left) or
MARE3 located upstream of mmyR (right) in response to increasing amounts of
AHFCA1. Lane 1: DNA fragment only. Lanes 2 to 7: AHFCA1 at 0, 0.8, 8, 20, 40 and
100 nmol, respectively. Amounts of DNA and proteins were kept constant (0.8 pmol and
4.0 pmol respectively).
3.2.2. Comparison of the binding affinity of MmfR to three MAREs
In order to further understand the role of MmfR in the regulation of methylenomycin
biosynthesis, determination of the relative binding affinities of MmfR to the three
operator regions was investigated. We thereby undertook competition assays, in which
two different MARE sequences were present in the same reaction, with increasing
concentrations of MmfR. In order to differentiate between the different sequences on a
gel, both long and short versions were prepared as shown in Table 3.1.
Table 3.1. Longer and shorter DNA sequences containing each of MAREs.
ARE No. Location of DNA Size of large (bp) Size of short (bp)
MARE1 mmfR-mmfL 194 100
MARE2 mmyY-mmyB 230 98
MARE3 upstream of mmyR 150 100
DNA fragments were amplified for each region by PCR using cosmid C73-787 as
template and primers listed in Table 6.1. For MARE1, located in the mmfR-mmfL
intergenic region, the entire intergenic region (194 bp) and a partial fragment (100 bp)
were amplified. For MARE2, located in the mmyY-mmyB intergenic region, the entire
intergenic region (230 bp) and a partial fragment (98 bp) were amplified. For MARE3
located upstream of mmyR, the entire mmfP-mmyR intergenic region (150 bp) and one
shorter fragment (100 bp) were prepared.
Chapter 3 : Interactions between DNA, ligand and protein
65
All the fragments still contain their specific MARE sequences. Agarose gel
electrophoresis was used to separate the desired PCR products from unused reactants and
template DNA (Figure 3.8). The PCR products were excised from the gel and
subsequently purified and sequenced.
Figure 3.8. Agarose gel electrophoresis analyses of the PCR products for different length
of DNA fragments containing MAREs. (LR: low range DNA ladder)
The two operators, one located between mmfR and mmfL and the other between mmyY
and mmyB, were compared in EMSAs on 6% non-denaturing polyacrylamide gels as
shown in Figure 3.9.
Figure 3.9. Competitive binding of MmfR with two DNA fragments corresponding to
the mmfR-mmfL and mmyY-mmyB intergenic regions. Lanes 1 to 3: DNA fragments only.
Lanes 4 to 9: MmfR at 0.1, 0.2, 0.3, 0.4, 0.5 and 0.6 pmol, respectively. The amount of
DNA was kept constant (0.05 pmol).
As shown in the left gel in Figure 4.9, the mmfR-mmfL fragment was larger than mmyY-
mmyB while in the right gel, the mmfR-mmfL fragment was shorter than mmyY-mmyB.
Chapter 3 : Interactions between DNA, ligand and protein
66
Figure 3.9 (left gel) shows the band shift corresponding to the DNA fragment between
mmyY-mmyB occurs at lower protein concentration than the band shift for the fragment
between mmfR and mmfL. To check this is not just a product of the different DNA sizes
used, the experiment was repeated with a larger mmyY-mmyB DNA fragment and a
smaller mmfR-mmfL DNA fragment (Figure 3.9, right gel). Since the same result was
observed, this strongly indicates that MmfR binds the operator between mmyY and mmyB
more strongly than that between mmfR and mmfL.
Using the same technique, binding of MmfR to the operator regions, located between
mmfR and mmfL and upstream of mmyR, was compared (Figure 3.10). The right gel with
a longer mmfR-mmfL intergenic DNA and shorter sequence upstream of mmyR, did not
give useful information, without obvious band shifts until the maximum MmfR
concentration. In the left gel, the shorter DNA fragment corresponding to the operator
between mmfR and mmfL, exhibited band shifts with lower concentrations of MmfR,
suggesting that the affinity of MmfR to this operator is higher than to the one upstream
of mmyR. However, it was noted that there may be some crosslinking of the DNA
fragments when the highest amount of MmfR (0.6 pmol) used, as shown in the rightmost
lane of the gel.
Figure 3.10. Competitive binding of MmfR with two DNA fragments corresponding to
the mmfR-mmfL intergenic region and upstream of mmyR. Lanes 1 to 3: DNA fragments
only. Lanes 4 to 9: MmfR at 0.1, 0.2, 0.3, 0.4, 0.5 and 0.6 pmol, respectively. The amount
of DNA was kept constant (0.05 pmol).
The above EMSAs provided an affinity order of MmfR to the three MAREs as followed:
the operator between mmyY and mmyB > the operator between mmfR and mmfL > the
operator upstream of mmyR.
In order to further confirm this result, the affinity comparison of MmfR to the operator
located between mmyY and mmyB and the operator upstream of mmyR was also
performed. However, the EMSAs did not work well and the band shifts seemed unclear.
Chapter 3 : Interactions between DNA, ligand and protein
67
From the gels shown in Figure 3.11, it can be presumably concluded from the band shifts
that the operator between mmyY and mmyB has higher affinity to MmfR compared to the
operator upstream of mmyR. This is consistent with the affinity order obtained from the
above two competitive binding assays. These results would provide more evidence for
understanding the role of MmfR in the regulation of methylenomycin biosynthesis in S.
coelicolor A3(2).
Figure 3.11. Competitive binding of MmfR with two DNA fragments corresponding to
the mmyY-mmyB intergenic region and upstream of mmyR. Lanes 1 to 3: DNA fragments
only. Lanes 4 to 9: MmfR at 0.1, 0.2, 0.3, 0.4, 0.5 and 0.6 pmol, respectively. The amount
of DNA was kept constant (0.05 pmol).
3.3 Interactions of MmfR with natural AHFCAs
As discussed in Section 1.2.1, there are five natural AHFCAs that act as autoinducers for
methylenomycin production in S. coelicolor. In order to assess their relative ability to
bind MmfR and release the protein from cognate operators, EMSAs were carried out on
6% non-denaturing polyacrylamide gels for each AHFCA compound using DNA
fragments of the entire mmfR-mmfL or the entire mmyY-mmyB intergenic region. These
fragments were chosen because they have the highest affinity to MmfR and their long
size provides better EMSA images after DNA staining and visualisation under UV.
3.3.1 Synthesis of AHFCA1 and AHFCA3
The ligands, AHFCA2, AHFCA4 and AHFCA5, used in the EMSAs were obtained from
Nicolas Malet. AHFCA1 and AHFCA3 were chemically synthesised in this study. The
synthesis of AHFCAs has been well established in the Challis group based on a reported
synthetic route by Sello and co-workers.76-77
Chapter 3 : Interactions between DNA, ligand and protein
68
Scheme 3.1. Synthesis of AHFCAs reported by Sello and co-workers.76
The reported synthesis of AHFCAs utilised scandium triflate as a Lewis acid catalyst for
the condensation of dihydroxyacetone acetonide 109 with a β-ketoester, affording methyl
2-alkyl-4-hydroxymethylfuran-3-carboxylates. Subsequent lithium hydroxide-mediated
hydrolysis of the methyl ester gave the desired AHFCAs (Scheme 3.1). This synthesis
was identified accidentally during their attempts to synthesise γ-butyrolactone signalling
molecules via butenolide intermediates.76 The scandium triflate-catalysed reaction has
been noted as a variant of the Garcia Gonzalez reaction (Scheme 3.2).
Scheme 3.2. General concept of the Garcia Gonzalez reaction.
A possible mechanism for the condensation of 109 and the β-ketoester has been proposed
involving scandium triflate-mediated in situ deprotection of 109 and enolisation of the
β-ketoester. Knoevenagel condensation between dihydroxyacetone and the enol,
followed by addition of the hydroxyl group to the ketone yielding a ketal intermediate
that undergoes dehydration and aromatisation to give the furan products (Scheme 3.3).77
Chapter 3 : Interactions between DNA, ligand and protein
69
Scheme 3.3. Proposed mechanism for Knoevenagel condensation and subsequent furan
formation catalysed by Sc(OTf)3.77
The synthetic approach is characterised by its low cost of catalyst, utilising only 10 mol%
of scandium triflate, and mild reaction conditions at room temperature in good yield. The
reactions are neither air- nor moisture-sensitive which makes it very easy to handle. This
route has previously been confirmed as a very reproducible way for synthesis of
AHFCAs in the Challis group. Thus it was used in this work for synthesis of AHFCA1
and AHFCA3, as well as AHFCA6 and AHFCA7 in Section 3.6.3.
The synthesis of the β-ketoester starting materials was accomplished via acyl Meldrum’s
acid intermediates, which were prepared from the corresponding acid or acid
chloride.77,167 Both 3-methylbutyryl chloride and 4-methylpentanoic acid are available in
our lab. The mechanism for the coupling of the carboxylic acid and Meldrum’s acid,
Chapter 3 : Interactions between DNA, ligand and protein
70
induced by EDC and DMAP is described in Section 2.2. Hydrolysis of the acyl
Meldrum’s acid in methanol gives the corresponding β-ketoester (Scheme 3.4).
Scheme 3.4. Synthesis of the β-ketoesters 110 and 112 for AHFAC1 and AHFCA3,
respectively.
Scandium triflate-catalysed condensation of dihydroxyacetone acetonide 109 with the β-
keto esters 110 and 112 afforded the corresponding furans 113 and 114 in good yield
after purification by flash chromatography. The ester group within these furans was
hydrolysed with lithium hydroxide, yielding the AHFCA1 26 and AHFCA3 42 as waxy
solids (Scheme 3.5). The overall yields for this two-step process are 44% for AHFCA1
and 37% for AHFCA3.
Scheme 3.5. Synthesis of AHFCA1 and AHFCA3.
3.3.2 Comparison of the binding affinity of the different intergenic regions to
MmfR with natural AHFCAs 1-5
The EMSAs showing the effect of increasing concentration of different AHFCAs on
binding of MmfR to the 194-bp DNA fragment of the entire intergenic region between
mmfR and mmfL is shown in Figure 3.12. The more DNA released from MmfR, the more
active the AHFCA is. Due to the semi-quantitative nature of EMSAs and the complicated
interactions in such a three-component system involving DNA, protein and ligand,
Chapter 3 : Interactions between DNA, ligand and protein
71
quantification of the DNA band released by addition of different AHFCAs is difficult.
However, some general conclusions can be made.
Undoubtedly, AHFCA1 was shown to be the strongest ligand among the five to interact
with MmfR and cause release of the DNA, while AHFCA2 obviously showed the lowest
amount of DNA release. AHFCA5 exhibited similar activity to AHFCA4 and the potency
of AHFCA3 seems higher than both of them but lower than AHFCA1. Thus, based on
the EMSAs using the 194-bp DNA fragment of the entire mmfR-mmfL intergenic region,
the order of the activity of the five natural AHFCAs was concluded to be as follows:
AHFCA1 > AHFCA3 > AHFCA5 ≈ AHFCA4 > AHFCA2.
Figure 3.12. Interaction of MmfR with the 194-bp DNA fragment of the entire mmfR-
mmfL intergenic region in response to increasing amounts of natural AHFCAs 1-5. Lane
1: DNA fragment only. Lanes 2 to 9: AHFCAs at 0, 0.8, 4, 8, 14, 20, 40 and 100 nmol,
respectively. Amounts of DNA and proteins were kept constant (0.1 pmol and 1.8 pmol
respectively).
The results with the 230-bp DNA fragment of the entire intergenic region between mmyY
and mmyB are shown in Figure 3.13. Compared to Figure 3.12, it can be clearly seen that
the ability of AHFCAs 1-5 to release the mmyY-mmyB intergenic DNA from the DNA-
MmfR complex is lower than the mmfR-mmfL intergenic DNA sequence. This indicates
that the binding affinity of MmfR to MARE2 located between mmyY and mmyB is
stronger than to MARE1 between mmfR and mmfL, further supporting the affinity order
confirmed in Section 3.2. Generally, AHFCA1 and AHFCA3 still showed higher efficacy
than the other three AHFCAs.
Chapter 3 : Interactions between DNA, ligand and protein
72
Figure 3.13. Interaction of MmfR with the 230-bp DNA fragment of the entire mmyY-
mmyB intergenic region in response to increasing amounts of natural AHFCAs 1-5. Lane
1: DNA fragment only. Lanes 2 to 9: AHFCAs at 0, 0.8, 4, 8, 14, 20, 40 and 100 nmol,
respectively. Amounts of DNA and proteins were kept constant (0.1 pmol and 1.8 pmol
respectively).
3.4 Key residues in the ligand binding of AHFCAs by MmfR
3.4.1 Analysis of the crystal structure
As discussed in Section 1.3, the TetR family transcriptional regulators are highly
conserved in their N-terminal DNA-binding domain while the rest of the sequence is less
conserved.78 This, to some extent, explains why these repressor proteins bind to ARE
consensus sequences. It is possible that minor alterations in the sequence of the HTH
motif formed by the N-terminal residues dictate their specificity for specific DNA
binding sites. Moreover, it is likely that the less conserved part in these receptor proteins
allows for variability in their ligand binding specificity. Thus, they are responsive to
certain structural types of ligands.
The MmfR structure with AHFCA2 bound is the first example of an ArpA-like repressor
structure with its ligand bound. This allows us to identify the ligand binding site in MmfR
as shown in Figure 3.14. From a detailed analysis of the structure, it appears that the
hydroxyl group on the phenyl ring of Tyr 85 participates directly in binding AHFCAs
through hydrogen bond formation with their hydrophilic part. Additional residues may
be involved in stabilising the interaction, such as Gln 130. The amide group at the end of
Chapter 3 : Interactions between DNA, ligand and protein
73
its side chain is in close proximity to the hydrophilic groups of AHFCA2. Although the
side chain of Tyr 144 is pointing away from the ligand, its backbone amide seems to be
involved in the interaction with the carboxylic acid group of the ligand. Additionally,
there is a pocket for the hydrocarbon chain of AHFCA2, in which a residue, Leu 110, is
situated at the bottom. The presence of Leu 110 at the bottom of the hydrophobic pocket
may determine the ligand specificity of MmfR to AHFCAs with alkyl chains of modest
length. It is tempting to speculate that a smaller residue, such as glycine, may contribute
to the formation of a deeper hydrophobic pocket for the alkyl chain of AHFCAs. As
shown in Figure 1.5, longer side chains are often seen in GBLs, such as SCBs.
Figure 3.14. (A) Proposed key residues of MmfR interacting with AHFCA2 from solved
crystal structure. (B) Key hydrogen bonding interactions in the hydrophilic part of
AHFCA2.
In order to probe the above hypothesis, site-directed mutagenesis studies, followed by
EMSAs, were carried out. Mutations were focused on the proposed key residues in
binding AHFCAs, including Tyr 85, Gln 130, Tyr 144 and Leu 110.
3.4.2 Preparation of MmfR mutants
Since the hydroxyl group of Tyr 85 was proposed to form a hydrogen bond with the
carboxylic acid group and the hydroxyl group of AHFCA ligand, Tyr 85 was mutated to
a phenylalanine, as well as an alanine. As the amide group on the side chain of Gln 130
was proposed to participate in the hydrogen bond formation with the hydrophilic part of
AHFCAs, Gln 130 was mutated to a glutamic acid, which does not contain an amide
group at the end of its side chain. From the crystal structure at the position of Tyr 144, it
was proposed that the backbone amide may be involved in the interaction with the
carboxylic acid group of AHFCAs. In order to confirm this, Tyr 144 was mutated to a
Chapter 3 : Interactions between DNA, ligand and protein
74
phenylalanine, as well as an alanine. The plasmids of MmfR mutants with alanine were
provided by Kathryn Styles, a PhD student in the Corre group.
The pET151/D-TOPO plasmid containing mmfR was used as the template for site-
directed mutagenesis using Q5 site-directed mutagenesis kit with primers and conditions
listed in Table 6.2. The constructs were sequence verified. E. coli BL21 star (DE3)
competent cells were then transformed with these recombinant plasmids. It should be
noted that E. coli BL21 star (DE3)pLysS competent cells were transformed with the
recombinant plasmid of mutant Y144F. This is due to a slow growth of E. coli cells,
around 6 hours to reach OD600 > 0.6, and low expression level of the protein when the
mutant Y144F was expressed in E. coli BL21 star (DE3) cells. E. coli BL21 star
(DE3)pLysS cells are more suitable for expression of potentially growth-inhibiting
recombinant proteins due to its expression of T7 lysozyme, a T7 RNA polymerase
inhibitor that prevents leaky expression in un-induced cells.
MmfR mutants were overproduced and purified as N-terminal His6-tagged proteins.
SDS-PAGE analysis confirmed they had the expected molecular weight (~27.8 kDa)
(Figure 3.15).
Figure 3.15. SDS-PAGE analysis of His6-MmfR mutants, Y85F, Y85A, Q130E, Y144F
and Y144A, after purification. (MWM: molecular weight marker)
The identity of these MmfR mutants, Y85F (27819.5 Da), Y85A (27743.5 Da), Q130E
(27836.5 Da), Y144F (27819.5 Da) and Y144A (27743.5 Da), was further confirmed by
LC-HRMS (Figure 3.16).
Chapter 3 : Interactions between DNA, ligand and protein
75
Figure 3.16. Measured (top) and deconvoluted (bottom) mass spectra of His6-MmfR
mutants, Y85F (calculated mass = 27819.5 Da), Y85A (calculated mass = 27743.5 Da),
Y144F (calculated mass = 27819.5 Da), Y144A (calculated mass = 27743.5 Da) and
Q130E (calculated mass = 27836.5 Da).
3.4.3 Attempts to mutate Leu 110
Considerable efforts were made to mutate Leu 110 to spatially smaller residues, such as
glycine, alanine and valine, thus leading to a deeper binding pocket probably suitable for
AHFCA ligands with longer alkyl chains.
Chapter 3 : Interactions between DNA, ligand and protein
76
The first mutation tried was the L110G mutant, which failed due to low expression of
the protein and difficult purification. Expression of the MmfR mutant L110A in E. coli
BL21 star (DE3) cells encountered problem of growth inhibition and very low level or
even no expression of the protein, similarly to the mutant Y144F, as well as purification
problems. The problem was not solved even when the mutant L110A was expressed in
E. coli BL21 star (DE3)pLysS cells. More surprisingly, the L110V mutant was found to
be present in the insoluble fraction as shown in Figure 3.17. The above results indicate
that probably Leu110 is vital for proper folding and stability of MmfR.
Figure 3.17. SDS-PAGE analysis in the attempt to purify the MmfR mutant L110V.
3.4.4 Effect of mutations on ligand binding
The activity of MmfR mutants, Y85F, Y85A, Q130E, Y144F and Y144A, was assessed
using EMSAs. The DNA fragment used was the 194-bp sequence of the entire mmfR-
mmfL intergenic region. The ligand used was AHFCA1, which showed the strongest
interaction with MmfR as described in Section 3.3.2.
To assure the mutants were still able to bind DNA, a pre-EMSA was performed for each
MmfR mutant in comparison with wild-type (WT) MmfR. With increasing
concentrations of protein, it showed that all the mutations of MmfR in this study do not
lead to altered DNA binding affinity for MmfR.
3.4.4.1. Tyr 85
The ability of MmfR mutants Y85F and Y85A and WT MmfR to bind DNA in the
presence of AHFCA1 were then compared (Figure 3.18).
Chapter 3 : Interactions between DNA, ligand and protein
77
Figure 3.18. Comparison of the effect of AHFCA1 on protein:DNA (mmfL-mmfR
fragment, 194 bp) dissociation when WT MmfR, MmfR Y85F and Y85A were used.
Lane 1: DNA fragment only. Lanes 2 to 9: AHFCA1 at 0, 0.8, 4, 8, 14, 20, 40 and 100
nmol, respectively. Amount of DNA and proteins were kept constant (0.1 pmol and 1.8
pmol respectively).
Both of the MmfR mutants at Tyr 85, Y85A and Y85F, showed significantly reduced
DNA release in the presence of AHFCA1 in comparison to WT MmfR. Interestingly, the
Y85F mutant, which only lacks the hydroxyl group on the side chain, gave a similar
result to the Y85A mutant. This highlights the importance of the hydroxyl group in the
interaction of MmfR with its ligand. The result is highly consistent with the information
about Tyr 85 obtained from the crystal structure.
3.4.4.2. Gln 130
The ability of MmfR mutant Q130E and WT MmfR to bind DNA in the presence of
AHFCA1 were compared (Figure 3.19).
Figure 3.19. Comparison of the effect of AHFCA1 on protein:DNA (mmfL-mmfR
fragment, 194 bp) dissociation when WT MmfR and MmfR Q130E mutant were used.
Lane 1: DNA fragment only. Lanes 2 to 9: AHFCA1 at 0, 0.8, 4, 8, 14, 20, 40 and 100
nmol, respectively. Amount of DNA and proteins were kept constant (0.1 pmol and 1.8
pmol respectively).
The Q130E mutant showed dramatically reduced ability to respond to AHFCA1 and
cause DNA release. Therefore, loss of the amide group of Gln 130 in this mutant
Chapter 3 : Interactions between DNA, ligand and protein
78
attenuates the ligand recognition and binding of AHFCAs by MmfR, which correlates
well with the proposed role for Gln 130 from the crystal structure.
3.4.4.3. Tyr 144
The ability of MmfR mutants Y144F and Y144A and WT MmfR to bind DNA in the
presence of AHFCA1 were compared (Figure 3.20).
Figure 3.20. Comparison of the effect of AHFCA1 on protein:DNA (mmfL-mmfR
fragment, 194 bp) dissociation when WT MmfR, MmfR Y144A and Y144F were used.
Lane 1: DNA fragment only. Lanes 2 to 9: AHFCA1 at 0, 0.8, 4, 8, 14, 20, 40 and 100
nmol, respectively. Amount of DNA and proteins were kept constant (0.1 pmol and 1.8
pmol respectively).
Figure 3.20 (right image) shows that AHFCA1 retained its ability to release DNA from
the mutant Y144F protein at a level that was comparable to, or possibly even more
effective than, WT MmfR (left image). This strongly indicates that it is not the side chain
of the residue Tyr 144 involved in the interaction with the AHFCA ligand. It is very
likely that the backbone amide at that position participates in ligand recognition and
binding with the hydrophilic portion of the AHFCAs, as proposed from the crystal
structure. However, no AHFCA1-induced release of the DNA was observed for the
mutant Y144A, probably due to alteration of the ligand binding pocket when Tyr 144
was replaced with a comparatively small residue.
3.5 Ligand binding specificity of MmfR
As discussed in Section 1.2, there are two structural types of signalling molecules present
in S. coelicolor, one of which is the γ-butyrolactone type SCBs, while the other is the
furan type AHFCAs for the regulation of methylenomycin biosynthesis. In order to
determine whether the regulation systems in S. coelicolor A3(2) are pleiotropic or
pathway-specific, we investigated the ability of MmfR to interact with SCB1. Thus, we
Chapter 3 : Interactions between DNA, ligand and protein
79
chemically synthesised the γ-butyrolactone signalling molecule SCB1 and assessed the
responsiveness of MmfR to it by EMSAs.
3.5.1 Synthesis of SCB1
The total synthesis of GBLs has been reported many times.168,173-178 As A-factor was the
first identified γ-butyrolactone signalling compound, most efforts have been made on its
synthesis. A most common strategy involves acylation of a protected β-hydroxymethyl-
γ-butyrolactone to introduce an α acyl group (Scheme 3.6). However, lots of evidence
suggested that preparation of the protected β-hydroxymethyl-γ-butyrolactone is not easy
and would be the key step in the route.
Scheme 3.6. Key steps in most reported synthesis of A-factor.
The structure of SCB1 15, (2R,3R,1'R)-2-(1'-hydroxy-6-methylheptyl)-3-hydroxymethyl
butanolide, was first determined by Takano and co-workers in 2000 by comparing the
purified metabolite with the chemically synthesised SCB1 and its stereoisomers.57 Their
synthetic route is shown in Scheme 3.7 by using the aforementioned strategy for
constructing the α-position acylated β-hydroxymethyl-γ-butyrolactone in the presence of
lithium diisopropylamide (LDA). Reduction of A-factor with NaBH4 and separation by
C18 HPLC gave the IM-2-type enantiomers (2R,3R,6R and 2S,3S,6S) and VB-type
enantiomers (2R,3R,6S and 2S,3S,6R) of SCB1. They further separated each racemic
mixture by chiral HPLC to yield optically pure compounds in order to characterise the
absolute configuration of natural SCB1, which has been determined as 2R,3R,6R.
Chapter 3 : Interactions between DNA, ligand and protein
80
Scheme 3.7. Synthesis of SCB1 reported by Takano and co-workers.57
More recently, an alternative synthetic methodology was reported by Sello and co-
workers for the synthesis of A-factor (Scheme 3.8).168 Compared to previous syntheses,
this approach provides a more convenient access to the core structure of GBLs using
inexpensive and easily handled starting materials and reagents. It was also used in
preparation of the protected butenolide for investigation of MmfL catalytic activity and
discussed in Section 2.3. The alkene within the resulting butenolide is reduced using
sodium cyanoborohydride and subsequent deprotection of the TBDMS group with acid
affords racemic A-factor. Thus, an efficient stereoselective synthesis of A-factor and
related GBLs has yet to be achieved.
Chapter 3 : Interactions between DNA, ligand and protein
81
Scheme 3.8. Synthesis of A-factor reported by Sello and co-workers.168
The structure of SCB1 is very close to A-factor except that SCBs possess a hydroxyl
group instead of a ketone in their alkyl chains. This feature results in the requirement of
an additional reduction step in their synthesis compared with A-factor. Thus the reported
route by Sello et al. was slightly modified to adapt for our specific target molecule. For
reduction of the alkene on the butenolide ring, Sello et al. used sodium cyanoborohydride
in ethanol as the reducing condition, with strict control of the reaction time to selectively
reduce the alkene without over-reduction of the ketone group on the side chain. In
contrast to their synthesis, reduction of the alkene and ketone was required for SCB1 and
so a more potent reducing agent was required. Selectivity for the ester was still required.
So sodium cyanoborohydride and acetic acid was used and provided the desired alcohol
after 8 hours in a yield of 66%.
For the purposes of optimisation, we first used heptanoic acid which is commercially
available instead of 6-methylheptanoic acid, therefore resulting in the preparation of a
SCB1 analogue 118, namely SCB6 (Scheme 3.9).60
Scheme 3.9. Synthesis of SCB6 for optimisation purpose.
Chapter 3 : Interactions between DNA, ligand and protein
82
6-Methylheptanoic acid was prepared using a synthetic route similar to that used by
Takano and co-workers. To simplify the preparation for acyl Meldrum’s acid 122, the
combination of EDC and DMAP was used for coupling of 6-methylheptanoic acid with
Meldrum’s acid. This is same as the synthesis of acyl Meldrum’s acid for NAC β-
ketothioesters described in Section 2.2 and preparation of AHFCAs described in Section
3.3.1. Our synthetic route for SCB1 is summarised in Scheme 3.10.
Scheme 3.10. Synthesis of SCB1 and its stereoisomers in this study.
There are three stereogenic centers formed after the reduction. Similar to that reported
by Takano and co-workers using NaBH4, it resulted in formation of two pairs of
enantiomers. The production of four stereoisomers instead of the possible eight is due to
the steric hindrance of the molecule, which only results in products with the two groups
at C-2 and C-3 position of the butyrolactone ring on opposite side. The products involve
the IM-2 type enantiomers (2R,3R,6R and 2S,3S,6S) and VB type enantiomers (2R,3R,6S
and 2S,3S,6R) of SCB1.
The separation of the two pairs of enantiomers were performed on column
chromatography with silica gel after deprotection of the TBDMS group to yield the
desired racemic SCB1. The 400 MHz 1H NMR spectra of racemic SCB1 and the other
VB type compounds are identical to the reported data, confirming that SCB1 was
Chapter 3 : Interactions between DNA, ligand and protein
83
successfully prepared. The difference between racemic SCB1 and their diastereoisomers
in 1H NMR is obvious (Figure 3.21). In addition, the spectra of SCB6 118 also showed
the same key features as SCB1. Given the qualitative purpose of our subsequent
bioassays, racemic SCB1 was not further separated and directly applied to the following
experiments.
Figure 3.21. Comparison of the 400 MHz 1H NMR spectra of racemic SCB1 and their
diastereoisomers.
3.5.2 Interaction of MmfR with SCB1
In order to assess whether MmfR is responsive to SCB1, EMSAs were performed with
the 194-bp DNA fragment of the entire intergenic region between mmfR and mmfL. The
EMSA result with increasing concentrations of racemic SCB1 was compared to that with
AHFCA1 (Figure 3.22).
In order to obtain reliable data, the concentration of racemic SCB1 was increased to
20 mM, which is four times higher than that previously used for other EMSAs. The DNA
and protein were prepared from the same batch and their concentrations were kept
constant.
The difference in ability to release MmfR from its operator is clear between SCB1 and
AHFCA1. Addition of AHFCA1 resulted in a band shift caused by the DNA release from
MmfR-DNA complex at a concentration of 0.4 mM or higher. However, the EMSA with
same amounts of SCB1 did not lead to a band shift. In fact, no DNA band shift can be
Chapter 3 : Interactions between DNA, ligand and protein
84
seen even when the concentration of SCB1 was increased to 20 mM. This indicates that
the γ-butyrolactone signalling molecule SCB1 could not bind to MmfR and initiate its
release from the operator. The unresponsiveness of MmfR to γ-butyrolactone implies
that its ligand binding is highly specific, supporting the hypothesis that GBLs and
AHFCAs mediated regulation is pathway-specific in S. coelicolor A3(2).
Figure 3.22. Interaction of MmfR with the 194-bp DNA fragment of the entire mmfR-
mmfL intergenic region in response to increasing amounts of AHFCA1 and racemic
SCB1. Lane 1: DNA fragment only. Lanes 2 to 7: AHFCA1 or racemic SCB1 at 0, 0.8,
8, 20, 40 and 400 nmol, respectively. Amounts of DNA and proteins were kept constant
(0.1 pmol and 1.6 pmol respectively).
3.6 Investigation of a MmfR homologue, SgnR
As discussed in Section 1.3, SgnR has been identified from the gbn cluster in S.
venezuelae as a homologue of MmfR with high sequence similarity. ARE sequences,
which are situated in the region upstream of gbnA, the intergenic regions of sgnR-sgnL
and gbnR-sgnH, are nearly identical to those identified in the S. coelicolor mmy gene
cluster (Figure 1.15 and Table 1.1). Moreover, AHFCA-like compounds, including
AHFCA6 and AHFCA7, were identified as the metabolites from the culture and
proposed to be the autoinducers for gaburedin production from the gbn cluster.60
The high similarity in the regulation system between the S. venezuelae gbn cluster and
the S. coelicolor mmy cluster intrigued us. In order to probe whether the regulatory
system involving SgnR in the gbn cluster shares the same features as MmfR in
methylenomycin production, we used the same methodology described previously to
study the interactions between SgnR, the AREs and the identified AHFCA compounds,
such as AHFCA6. The DNA-binding ability of SgnR for each of the AREs as well as
their interaction with AHFCA6 was investigated in vitro by EMSAs.
Chapter 3 : Interactions between DNA, ligand and protein
85
3.6.1 Cloning of sgnR
The sgnR gene sequence was amplified by PCR from the SV4 F01 cosmid using primers
designed to ensure directional cloning of the PCR products into pET151/D-TOPO vector
as described in Section 6.3.13. Agarose gel electrophoresis was used to separate the
desired PCR products from unused reactants and template DNA (Figure 3.23).
Figure 3.23. Agarose gel electrophoresis analysis of the sgnR PCR products (604 bp).
(LR: low range DNA ladder)
The PCR product was excised from the gel and subsequently purified. The purified PCR
product was then cloned into pET151/D-TOPO and E. coli TOP10 cells (IBA StarGate)
were transformed with the new plasmid. Positive colonies containing the new plasmid
pET151-sgnR were identified by selection on ampicillin plates, followed by restriction
digest with PstI to give two fragment bands at 5760 and 1585 bp (Figure 3.24, left image).
Insertion of the correct DNA sequence into the plasmid was also checked by PCR
amplification of the gene of interest using the same primer pairs (Figure 3.24, right
image).
Figure 3.24. Agarose gel electrophoresis analysis of restriction digests (left) using PstI
to confirm the constructed pET151-sgnR plasmid, and PCRs (right) using the constructed
pET151-sgnR plasmid as template (604 bp). (LR: low range DNA ladder)
Chapter 3 : Interactions between DNA, ligand and protein
86
Plasmids that gave both PCR products of the expected size and restriction digest patterns
consistent with the insertion of the genes of interest were sent for sequencing, with
amplification using the T7 primer pair. Sequencing data confirmed that the plasmid
contained the intact sequence of sgnR.
3.6.2 Overproduction and purification of recombinant His6-SgnR protein
Chemically-competent E. coli BL21 star (DE3) cells were transformed with the pET151-
sgnR plasmid. Total and soluble protein samples from the IPTG-induced expression of
each gene were analysed by SDS-PAGE. The SgnR protein was overproduced and a clear
band at the expected size (~26.2 kDa) was present in the total protein sample and the
soluble fraction (Figure 3.25, left image). His6-SgnR was purified by nickel affinity
chromatography as described in Section 6.3.16. SDS-PAGE analysis of the purified His6-
SgnR protein confirmed it had the expected molecular weight (Figure 3.25, right image).
Figure 3.25. SDS-PAGE analysis of total (PTOT) and soluble (PSol) protein fractions
resulting from overproduction of His6-SgnR in E. coli BL21star (DE3) (left), and His6-
SgnR after purification (right). (MWM: molecular weight marker)
The identity of the purified recombinant SgnR was further confirmed by LC-HRMS. The
determined molecular weight is consistent with the expected for His6-SgnR (26159.5 Da).
(Figure 3.26).
Chapter 3 : Interactions between DNA, ligand and protein
87
Figure 3.26. Measured (top) and deconvoluted (bottom) mass spectra of His6-SgnR
(calculated mass = 26159.5 Da).
3.6.3 Synthesis of AHFCA6 and AHFCA7
AHFCA6 and AHFCA7 were chemically synthesised using the same synthetic
methodology as for AHFCA1 and AHFCA3 described in Section 3.3.1. As the alkyl
chains in AHFCA6 and AHFCA7 were an isohexyl and a hexyl group, respectively, the
starting acylation material for the acyl Meldrum’s acid was 5-methylhexanoic acid for
AHFCA6 and heptanoic acid for AHFCA7.
The combination of EDC and DMAP was used for the coupling between the carboxylic
acid with Meldrum’s acid. After formation of the β-ketoesters 127 and 128, they were
separately condensed with dihydroxyacetone acetonide 109 in the presence of scandium
triflate to afford the corresponding furans 129 and 130 after purification by flash
chromatography. The ester group within these furans was hydrolysed with lithium
hydroxide, yielding the AHFCA6 45 and AHFCA7 46 as waxy solids. A summary of the
synthesis of AHFCA6 and AHFCA7 is described in Scheme 3.11.
Chapter 3 : Interactions between DNA, ligand and protein
88
Scheme 3.11. Synthesis of AHFCA6 and AHFCA7.
The prepared AHFCA6 and AHFCA7 were used as synthetic standards in LC-MS
comparison with the S. venezuelae metabolites for their structure elucidation (work by J.
Sidda).60
3.6.4 EMSAs with SgnR
Due to the high sequence similarity of AREs present in the gbn cluster with those in the
methylenomycin biosynthetic gene cluster (Table 1.1), the DNA-binding affinity of
SgnR was first tested in EMSAs with the 194-bp DNA fragment of the entire mmfR-
mmfL intergenic region on 6% non-denaturing polyacrylamide gel. A complete band shift
was observed when the concentration of SgnR was increased to 1.8 pmol, suggesting it
is capable of binding to the ARE sequence (Figure 3.27).
Figure 3.27. Interaction of SgnR with the 194-bp DNA fragment of the entire mmfR-
mmfL intergenic region. Lane 1: DNA fragment only. Lanes 2 and 3: SgnR at 0.9 and 1.8
pmol, respectively. Amount of DNA was kept constant (0.1 pmol).
The entire intergenic sequence between sgnR and sgnL (121 bp) was then amplified by
PCR from the cosmid SV4 F01 using primers listed in Table 6.1. The DNA fragments
Chapter 3 : Interactions between DNA, ligand and protein
89
containing the identified ARE1 sequence were analysed, purified from agarose gel
(Figure 3.28) and sent for sequencing. The correct DNA fragments confirmed by
sequencing were used in the EMSAs with SgnR on 6% non-denaturing polyacrylamide
gel.
Figure 3.28. Agarose gel electrophoresis analysis of the PCR products of the entire sgnR-
sgnL intergenic region (121 bp). (LR: low range DNA ladder)
The EMSA using the entire sgnR-sgnL intergenic DNA fragment containing the ARE1
sequence showed band shifts with addition of same concentrations of DNA and SgnR for
the mmfR-mmfL intergenic region (0.1 and 1.8 pmol, respectively) (Figure 3.29).
Moreover, the DNA fragments were released from the SgnR-DNA complexes in the
presence of increasing amount of AHFCA6. This strongly indicates the ability of SgnR
to bind the proposed operator between sgnR and sgnL and respond to AHFCA
compounds.
Figure 3.29. Interaction of SgnR with the 121-bp DNA fragment of the entire sgnR-sgnL
intergenic region in response to increasing amounts of AHFCA6. Lane 1: DNA fragment
only. Lanes 2 to 7: AHFCA6 at 0, 0.8, 8, 20, 40 and 400 nmol, respectively. Amounts of
DNA and proteins were kept constant (0.1 pmol and 1.8 pmol, respectively).
The in vitro interaction of SgnR with the other two operators, one located upstream of
gbnA and another between gbnR and sgnH, were also assessed with EMSAs with 10%
non-denaturing polyacrylamide gels (Figure 3.30). In order to test the DNA binding
Chapter 3 : Interactions between DNA, ligand and protein
90
ability of SgnR more specifically, the DNA fragments used were short annealed double-
stranded oligonucleotides (28 bp) containing the corresponding 18-bp ARE sequences.
The specific oligonucleotides are listed in Table 6.4 in Chapter 6. It can be seen that
SgnR could also bind to the other two operators and release the DNA in the presence of
AHFCA6, providing increased understanding of the regulatory mechanism involving
SgnR in the gbn cluter.
Figure 3.30. Interaction of SgnR with the 28-bp annealed DNA fragments containing
ARE2 located upstream of gbnA (left) or ARE3 located in the gbnR-sgnH intergenic
region (right) in response to increasing amounts of AHFCA6. Lane 1: DNA fragment
only. Lanes 2 to 7: AHFCA6 at 0, 0.8, 8, 20, 40 and 400 nmol, respectively. Amounts of
DNA and proteins were kept constant (0.8 pmol and 4.0 pmol respectively).
3.7 Key notes for studying in vitro DNA-protein-ligand interaction using EMSAs
From their extensive use in this study, EMSAs have proved to be an efficient way to
study the in vitro interaction not only between DNA and protein but also their interaction
with cognate ligands. Some general rules can be drawn from these studies as follows.
Firstly, the appropriate acrylamide concentration of the gel is important for running an
effective EMSA, which mainly depends on the size of oligonucleotides. It can be varied,
generally in the range of 5% to 25%. Lower percentage gels are better for bigger size
DNA fragments, while higher percentages are needed for smaller fragments, but resulting
in longer running time for a good image. We have learnt that 6% gel is enough for
EMSAs with DNA fragments longer than 90 bp and 10% gel better for oligonucleotides
less than 30 bp, such as the hairpin sequences and annealed 28-bp oligonucleotides used
in this study.
Moreover, the ratio of DNA to protein should be optimised in assays. This could be easily
determined by increasing the concentration of protein used for binding a certain amount
of DNA fragments. It is necessary to determine the amount of protein that completely
Chapter 3 : Interactions between DNA, ligand and protein
91
binds the amount of DNA used whilst not having too great excess. This is especially
important if a ligand compound will be added to test its ability to interact with the DNA-
protein complex and release DNA. Presence of “excessive” protein results in
consumption of more ligand to release the DNA. This is vital for EMSAs aiming at
comparison of ligand affinity and thus amounts of DNA and protein should be kept
constant in assays with each ligand. In an extreme case using too much protein, much
more ligand would be needed, thereby leading to precipitation of the protein. This results
in a “pseudo”-positive result that the compound used is effective in releasing DNA due
to no retardation by inactive protein, which would be misleading. However, this can be
easily distinguished from other samples during EMSAs due to its transformation to a
cloudy mixture after addition of the ligand as show in Figure 3.31.
Figure 3.31. An example of protein-DNA-ligand reaction with precipitation of the
protein in presence of the compound at a final concentration of 200 mM.
In addition, the DNA concentration should be added at the minimum level that allows
clear visualisation of the DNA band under UV after GelRed staining. Higher amount of
DNA leads to the use of more protein, thus requiring more ligand for DNA release. In
the case of DNA fragments longer than 90 bp, it has been established that the amounts
of DNA and protein for a good image are 0.1 pmol and 1.8 pmol, respectively. In the
case of DNA fragments shorter than 30 bp, 0.8 pmol of DNA and 4.0 pmol of protein
seems better. For DNA in the range of 30 to 90 bp, the amount and ratio of DNA and
protein still needs to be determined due to lack of relevant experience in this study.
Chapter 3 : Interactions between DNA, ligand and protein
92
3.8 Summary
From in vitro EMSA assays, it has been confirmed that three conserved MARE
sequences within the mmy gene cluster are the binding sites for MmfR, including the
operator upstream of mmyR and the operators located in the mmyY-mmyB, mmfR-mmfL
intergenic regions. The affinity order of MmfR to the three operators was demonstrated
from the competitive binding assays as followed: the operator between mmyY and mmyB >
the operator between mmfR and mmfL > the operator upstream of mmyR. This provides
more information about the regulation mechanism involving MmfR in methylenomycin
biosynthesis. Meanwhile, comparison of the ability of the five natural AHFCAs 1-5 on
the dissociation of the MmfR-DNA complexes showed their order as: AHFCA1 >
AHFCA3 > AHFCA5 ≈ AHFCA4 > AHFCA2.
Based on the X-ray crystal structure of MmfR-AHFCA2 complex, MmfR mutants at
specific residues, Tyr85Ala, Tyr85Phe, Tyr144Ala, Tyr144Phe and Gln130Glu, were
prepared using site-directed mutagenesis. In vitro interactions with EMSAs between
DNA, AHFCA1 and these mutants have revealed the molecular basis for AHFCA
recognition by MmfR. The hydroxyl group of Tyr85 and the amide group on the side
chain of Gln130 were confirmed to play important roles in the interaction of MmfR with
the ligand. Loss of these groups attenuates the ability of ligand to release DNA from
MmfR. For Tyr144, it is likely that its backbone amide participates in ligand recognition
and binding while its side chain may also be involved in the formation of the ligand
binding pocket. Leu110 is probably vital for proper folding and stability of MmfR.
SCB1, a γ-butyrolactone signaling molecule from S. coelicolor, has been successfully
prepared in this study. Current in vitro data obtained from EMSAs showed that SCB1
was unable to release MmfR from its operator. This demonstrates that ligand binding for
MmfR is highly specific, supporting the hypothesis that regulation of signalling system
with GBLs or AHFCAs is pathway-specific in S. coelicolor A3(2).
From the in vitro experiments with SgnR, it can be indicated that this MmfR homologue
is able to bind both the MARE sequence from S. coelicolor and the three ARE sequences
from S. venezuelae gbn gene cluster, including the operator upstream of gbnA and the
operators located in the sgnR-sgnL, gbnR-sgnH intergenic regions. The SgnR-DNA
complexes can be dissociated by addition of AHFCA6 to release the DNA on EMSAs,
Chapter 3 : Interactions between DNA, ligand and protein
93
indicating a similar regulation system as MmfR involved in the mmy gene cluster. Some
key notes for investigating the DNA-protein-ligand interactions on EMSAs have also
been summarised.
Chapter 4 : Watasemycin biosynthesis
94
4. Watasemycin biosynthesis in Streptomyces venezuelae
ATCC10712
Chapter 4 : Watasemycin biosynthesis
95
4.1 Introduction to previous work
4.1.1 Upregulation of ‘pyochelin-like’ gene cluster
BldM is an orphan atypical response regulator that plays a crucial role in morphological
differentiation.5 It was noted that transcription of a cluster containing fifteen genes,
sven0503-sven0517, is co-ordinately up-regulated in the S. venezuelae bldM mutant
compared to the wild type strain based on DNA microarray analyses (Figure 4.1).136
Additionally, Bibb et al. reported that transcription of the chloramphenicol biosynthetic
gene cluster is significantly increased in the same bldM mutant.179
Figure 4.1. Microarray expression profiles of the cluster of genes, sven0503-sven0517,
in S. venezuelae. The y axis shows normalised transcript abundance.136
Sequence analysis of the sven0503-sven0517 gene cluster revealed that many genes in
the cluster encode close homologues of the pyochelin biosynthetic enzymes in S. scabies
(Figure 4.2 and Table 4.1). A comparison of pyochelin biosynthesis in the Gram-negative
P. aeruginosa and the Gram-positive S. scabies can be found in Section 1.5.2.
Chapter 4 : Watasemycin biosynthesis
96
Figure 4.2. Organisation of the pyochelin and watasemycin biosynthetic gene clusters in
S. scabies 87-22 and S. venezuelae ATCC10712, respectively. Genes that encode
functionally analogous proteins are annotated in the same colour.
Sven0506 exhibits 82% sequence similarity to SCAB1381 (PchA/B) and is thus
proposed to be a salicylate synthase. Sven0510 is highly homologous to SCAB1411
(PchD), which is the A domain responsible for salicylate activation in S. scabies, and
Sven0511 is predicted to be a proofreading type II TE due to its sequence similarity to
SCAB1421 (PchC). Sven0512 and Sven0517 are homologues of the NRPSs SCAB1481
(PchE) and SCAB1471 (PchF), respectively. It is noteworthy that Sven0512 appears to
lack an MT-like domain, which is believed to catalyse the epimerisation of the
stereocenter in the first thiazoline ring during pyochelin biosynthesis in P. aeruginosa
and S. scabies. Both Sven0508 and Sven0516 show sequence similarity to PchK, which
has been proposed to catalyse reduction of the second thiazoline ring in the P. protegens
enantio-pyochelin biosynthetic pathway. There is one homologue of the AfsR-family
regulator SCAB1371 (Sven0505) and two homologues of the TetR-family regulator
SCAB1401 (Sven0507 and Sven0509). Finally, sven0515 putatively encodes a class B
RS methylase. As discussed in Section 1.6.1, these types of RS methylases are able to
catalyse methylation reactions on un-activated carbon centres. This indicates that a C-
methylation step may be involved in the biosynthesis of a pyochelin-like metabolite from
the sven0503-sven0517 gene cluster.
Chapter 4 : Watasemycin biosynthesis
97
Table 4.1. Proposed functions of proteins encoded by the watasemycin biosynthetic gene
cluster.136
Gene (accession
number)
Size
(bp) Homologue (accession number); origin
Identity/
similarity (%) Proposed function
sven0503
(CCA53790) 1250 SCAB1361 (CBG67360); S. scabies 87.22 44/51 AMP-ligase
sven0504
(CCA53791) 1247
Na+/H+ antiporter (WP_032767693);
Streptomyces sp. CNS654 88/91 Na+/H+ antiporter
sven0505
(CCA53792) 1947 SCAB1371 (CBG67361); S. scabies 87.22 59/80
AfsR-family
regulator
sven0506
(CCA53793) 1365 SCAB1381 (CBG67362); S. scabies 87.22 61/82 Salicylate synthase
sven0507
(CCA53794) 690 SCAB1401 (CBG67364); S. scabies 87.22 57/74
TetR-family
regulator
sven0508
(CCA53795) 1212
Trans-acting enoyl reductase (AEV87025);
Actinoplanes sp. SE50/110 36/46 Reductase
sven0509
(CCA53796) 687 SCAB1401 (CBG67364); S. scabies 87.22 70/88
TetR-family
regulator
sven0510
(CCA53797) 1602 SCAB1411 (CBG67365); S. scabies 87.22 71/79 Salicyl-AMP-ligase
sven0511
(CCA53798) 801 SCAB1421 (CBG67366); S. scabies 87.22 55/63 Thioesterase
sven0512
(CCA53799) 4545 SCAB1481 (CBG67372); S. scabies 87.22 59/68 NRPS
sven0513
(CCA53800) 1710 SCAB1431 (CBG67367); S. scabies 87.22 69/82
ABC transporter
permease/ATPase
sven0514
(CCA53801) 1842 SCAB1441 (CBG67368); S. scabies 87.22 68/79
ABC transporter
permease/ATPase
sven0515
(CCA53802) 1605
Radical SAM domain protein (AEV87031);
Actinoplanes sp. SE50/110 65/77
Radical SAM
methyltransferase
sven0516
(CCA53803) 1227
Trans-acting enoyl reductase (AEV87025);
Actinoplanes sp. SE50/110 48/56 Reductase
sven0517
(CCA53804) 5481 SCAB1471 (CBG67371); S. scabies 87.22 47/59 NRPS
4.1.2 Isolation and characterisation of watasemycin and thiazostatin from S.
venezuelae
Dr. Yuki Inahashi, a Postdoc in the Challis group, successfully expressed the sven0503-
sven0517 gene cluster in a Streptomyces superhost S. coelicolor M1152 and carried out
further LC-MS analyses. Streptomyces M1152 has been engineered to remove the
actinorhodin (act), prodiginine (red), calcium-dependent antibiotic (cda) and coelimycin
(cpk) gene clusters and contains a point mutation in rpoB, which encodes the RNA
polymerase β-subunit, to pleiotropically increase the production of specialised
metabolites.42
Chapter 4 : Watasemycin biosynthesis
98
LC-MS analyses of the culture supernatant of the S. coelicolor M1152/SV-2_E03::SspI
strain, which contains the sven0503-sven0517 gene cluster, identified metabolites giving
rise to ions with m/z corresponding to [M+H]+ for the known 2-hydroxyphenylthiazolines
thiazostatin 82, watasemycin 83, pyochelin 74, aerugine 75 and pulicatins A/B 77/78,
respectively, in comparison to wild type S. coelicolor M1152 (Figure 4.3). This was
further confirmed by LC-HRMS analysis carried out in this study (Figure 4.4).
Figure 4.3. Proposed structures of 2-hydroxyphenylthiazolines produced from the
sven0503-sven0517 gene cluster in S. venezuelae ATCC10712 based on high resolution
LC-MS analyses.
Chapter 4 : Watasemycin biosynthesis
99
Figure 4.4. Extracted ion chromatograms (EICs) at m/z = 353.0988, 339.0832, 210.0583,
224.0740 and 325.0675, corresponding to [M+H]+ for thiazostatin, watasemycin,
aerugine, pulicatins A/B and pyochelin, respectively, from LC-HRMS analyses of the
ethanol extract of S. coelicolor M1152/SV-2_E03::SspI culture supernatant (blue traces),
in comparison to wild type S. coelicolor M1152 (red traces). Further work in this study
would give more information about the peak corresponding to [M+H]+ for pyochelin (see
Section 4.2).
The main metabolic products with molecular formulae corresponding to thiazostatin and
watasemycin were extracted and isolated from the culture supernatant of S. coelicolor
M1152/SV-2_E03::SspI strain by Dr. Yuki Inahashi. NMR spectroscopic analysis
confirmed their identity by comparison to the reported data for thiazostatins A and B 82,
and watasemycins A and B 83, isolated from S. tolurosus and Streptomyces sp. TP-
A0597, respectively.116-117 Both thiazostatin and watasemycin are close analogues of
pyochelin 74 and enantio-pyochelin 76. Thiazostatin differs only by the addition of a
methyl group at C-4" of the thiazolidine ring while watasemycin, besides this, has an
additional methyl group at C-5' on the thiazoline ring. Other 2-hydroxyphenylthiazolines
putatively identified by LC-MS are pyochelin, aerugine and pulicatins A/B.
In order to investigate the biosynthesis of thiazostatin and watasemycin, characterisation
of the metabolite with molecular formula identical to pyochelin was pursued. As the
structure and biosynthesis of pyochelin 74 and enantio-pyochelin 76 have been well
Chapter 4 : Watasemycin biosynthesis
100
studied and characterised and are potential metabolites in biosynthesis, they would serve
as a model for investigating these pyochelin-like compounds. Section 4.2 will focus on
synthesis and characterisation of the metabolite with same molecular formula as
pyochelin.
As mentioned above, two putative thiazoline reductases, Sven0508 and Sven0516, are
present in the gene cluster, which is different from the P. aeruginosa or S. scabies
pyochelin biosynthetic pathway. Specific roles of these two enzymes in watasemycin
biosynthesis will be described in Section 4.3. Moreover, the absolute stereochemistry of
thiazostatin and watasemycin has not been previously characterised, with only relative
stereochemistry assigned. Section 4.4 will focus on reassignment of the relative and
absolute configuration of these compounds based on the obtained experimental results
and analysis. Section 4.5 will focus on studying the function of Sven0515, a putative
class B RS methylase, and the stereochemical course of the methylation reaction it
catalyses. A proposed biosynthetic pathway for watasemycin will be described in Section
4.6.
4.2 Identification and characterisation of isopyochelin
4.2.1 Synthesis of pyochelin
Assuming that the metabolite with molecular formula identical to pyochelin was actually
pyochelin, we aimed to chemically synthesise pyochelin and its stereoisomer
neopyochelin based on literature procedures.180-182 Then these compounds would be used
to compare to the pyochelin-like metabolite from S. coelicolor M1152/SV-2_E03::SspI
culture supernatant using LC-MS.
The first total synthesis of pyochelin was achieved in a three-step procedure by
Ankenbauer et al. in 1988 with a total yield of around 8% as shown in Scheme 4.1.180
Carboxylic acid 131 was prepared from 2-hydroxybenzonitrile and L-cysteine
hydrochloride in high yield. However, in their attempts to synthesise aldehyde 132, a
variety of reagents did not work except reduction of the carboxylic acid 131 by
thexylborane. Although the aldehyde 132 was produced, the yield of this step was very
poor (15%). The low yield was attributed to the major product being the alcohol resulting
from over-reduction of the aldehyde. Moreover, the aldehyde 132 was found to be
Chapter 4 : Watasemycin biosynthesis
101
exceedingly labile. Condensation of the aldehyde 132 and N-methyl-L-cysteine
hydrochloride was the final step to provide three stereoisomers of pyochelin: pyochelin
I, II and neopyochelin, with another stereoisomer not mentioned at that time.
Scheme 4.1. First total synthesis of pyochelin reported by Ankenbauer et al.180
The synthetic route was later modified by Rinehart et al. with successful assignment of
the absolute stereochemistry of the pyochelins.181 The reduction step to yield the
aldehyde 132 was reported to be significantly improved by using diisobutylaluminium
hydride (DIBAL-H) for reduction of the methyl ester 133 at -78 °C for 1 hour (Scheme
4.2). Again, the aldehyde 132 proved too unstable to provide acceptable analytical data.
The other steps were all same as that reported in the first total synthesis. The aldehyde
was immediately used for the condensation with N-methyl-L-cysteine to give four of the
eight possible pyochelin stereoisomers: pyochelins I and II 74, and neopyochelins I and
II (Figure 4.5). It was reported that the diastereomeric pyochelins cannot be separated in
pure form due to their rapid interconversion. Thus, assignment of their stereochemistry
was mainly based on spectral data of pyochelin methyl esters, which are comparably
more stable and can be purified, and the X-ray structure of a mutasynthetic pyochelin
analogue methyl ester.
Scheme 4.2. Improved synthesis of the aldehyde 132 by Rinehart et al.181
Chapter 4 : Watasemycin biosynthesis
102
Figure 4.5. The structure of synthetic pyochelin and neopyochelin.
To further improve the synthesis of pyochelin, Zamri and co-workers elected to use
alternative reaction conditions for the first step, i.e. condensation of
2-hydroxybenzonitrile with L-cysteine.182 The reaction was performed in a 1:1 mixture
of phosphate buffer 0.1 M, pH 6.4 and methanol at 50 °C for 4 days, in which
epimerisation of the carboxylic acid 131 at C-4' due to the very basic pH conditions
caused by piperidine could be avoided. They also used lithium aluminium hydride for
reduction of N-methoxy-N-methyl hydroxamate 134 to prepare aldehyde 132 (Scheme
4.3). The hydroxamate 134 was prepared using diethyl cyanophosphonate (DECP) as the
coupling agent for acid 131 and N,O-dimethylhydroxylamine in a yield of more than
90%. This synthetic approach represents the most efficient synthesis of pyochelin to date,
and has been used as the basis for the synthesis of pyochelin and analogues in our study.
However, the coupling agent DECP they used for preparation of hydroxamate 134 is
known to be highly toxic and moisture-sensitive and therefore required careful handling.
Scheme 4.3. Improved synthesis of aldehyde 132 by Zamri and co-workers.182
Our synthetic route for pyochelin was slightly modified from that reported by Zamri et al.
using 2-chloro-4,6-dimethoxy-1,3,5-triazine (CDMT) to couple acid 131 and N,O-
dimethylhydroxylamine, in the presence of N-methylmorpholine (NMM), as shown in
Scheme 4.4.
Chapter 4 : Watasemycin biosynthesis
103
Scheme 4.4. Synthetic route to pyochelin used in this study.
CDMT is a commercially available triazine-based coupling reagent, which works by
converting carboxylic acids to activated triazine esters. It requires the use of a tertiary
amine base, such as NMM, for acid activation, and the coupling can be carried out in
solvents such as THF, DMF, and EtOAc. The mechanism of its participation in coupling
reactions remains unclear. However, activation of the carboxylic acid via activated ester
or via acyl chloride have been proposed (Scheme 4.5).183 In situ NMM and CDMT form
the intermediate 4-(4,6-dimethoxy-1,3,5-triazin-2-yl)-4-methylmorpholinium chloride
(DMTMM), which can also be isolated and used as a coupling reagent independently.184
In each cycle of catalysis, 1-hydroxy-3,5-dimethoxytriazine is generated as a by-product.
CDMT coupling afforded N-methoxy-N-methyl hydroxamate 134 in a yield of 42%. The
modest yield seems to be due to generation of several by-products which made the
product purification by column chromatography difficult.
Chapter 4 : Watasemycin biosynthesis
104
Scheme 4.5. The proposed mechanism for CDMT-mediated coupling of acid 131 and
N,O-dimethylhydroxylamine.
Other steps used in the synthesis were same as those reported by Zamri et al.182 The
present synthesis gives only four stereoisomers instead of the possible eight allowed by
the three chiral centres, C-4', C-2" and C-4", in pyochelin. Since there is no report of
isomerisation at C-4" during thiazolidine syntheses under the present conditions, the
fixed centre, C-4", is assigned as R configuration, since it derives from N-methyl-L-
cysteine. The observed four diastereoisomers are thus due to isomerism at C-4' and C-2"
under the reaction conditions. The fact that the configuration at C-4" in pyochelin and
the α-carbon of N-methyl-L-cysteine used in the condensation are identical not only
benefits assignment of the pyochelin stereochemistry, as described by Rinehart et al., but
also aids in the study of other compounds of similar structural types.
4.2.2 LC-MS analysis of synthetic pyochelin
Due to the rapid interconversion of the pyochelin acids, synthetic product was used to
compare to the natural metabolites by LC-MS without further purification. Two of the
synthetic stereoisomers, pyochelins I and II, correspond to the two naturally occurring
Chapter 4 : Watasemycin biosynthesis
105
pyochelins 74 found in S. scabies and P. aeruginosa. However, none of these four
diastereomeric compounds had the same retention time as the metabolite with molecular
formula corresponding to pyochelin (Figure 4.6). The other four diastereoisomers of the
possible eight which are S configuration at C-4" would be enantiomers of these four
diastereoisomers we obtained, thus they would have the same retention times. Therefore,
it can be concluded that the metabolite is neither pyochelin nor any of its stereoisomers.
Figure 4.6. EICs at m/z = 325.0675, corresponding to [M+H]+ for the metabolite (top
trace) from LC-HRMS analyses of the ethanol extract of S. coelicolor
M1152/SV-2_E03::SspI culture supernatant, and synthetic mixture of pyochelins and
neopyochelins (bottom trace), including two stereoisomers identical to natural pyochelin
from S. scabies 87-22 and P. aeruginosa.
4.2.3 Synthesis and characterisation of isopyochelin
Considering both thiazostatin and watasemycin have a methyl group at C-4", we
hypothesised that the metabolite was a structural isomer of pyochelin 74 in which C-4"
is methylated instead of the nitrogen atom of the thiazolidine. To test this hypothesis, an
authentic standard of the C-4" methylated isomer of pyochelin, named as isopyochelin,
was synthesised from condensation of the aforementioned aldehyde 132 with 2-methyl-
L-cysteine, which cannot epimerise during the synthesis (Scheme 4.6).
Chapter 4 : Watasemycin biosynthesis
106
Scheme 4.6. Synthesis of a mixture of isopyochelin stereoisomers 135 with R
configuration at C-4".
A mixture of four synthetic diastereomeric compounds 135 with 4"R configuration were
compared with the metabolites with m/z of [M+H]+ = 325.0675 in the culture extract.
Two of the synthetic stereoisomers had the same retention time as the metabolites in the
extract (Figure 4.7), thus showing isopyochelin is a metabolic product of the gene cluster.
Figure 4.7. EICs for m/z = 325.0675, corresponding to [M+H]+ for isopyochelin, from
LC-HRMS analyses of the ethanol extract of S. coelicolor M1152/SV-2_E03::SspI
culture supernatant (top trace), the authentic synthetic standard of the isopyochelin
stereoisomers from 2-methyl-L-cysteine (middle trace) and the extract to which an
approximately equimolar quantity of the synthetic standard has been added (bottom
trace).
In order to assign the absolute configuration of isopyochelin, we further synthesised the
four stereoisomers resulting from condensation of aldehyde 132 with 2-methyl-D-
cysteine. This yielded isopyochelin stereoisomers 136 with 4"S configuration (Scheme
4.7).
Chapter 4 : Watasemycin biosynthesis
107
Scheme 4.7. Synthesis of a mixture of isopyochelin stereoisomers 136 with S
configuration at C-4".
The enantiopurity of the two commercially-sourced amino acids, 2-methyl-L-cysteine
and 2-methyl-D-cysteine, was confirmed by CD studies prior to commencing the
synthesis. Solutions containing an equimolar quantity of each amino acid showed mirror
image CD spectra, confirming their opposite absolute stereochemistry (Figure 4.8).
Figure 4.8. CD spectra of commercially available 2-methyl-L-cysteine and 2-methyl-D-
cysteine used in the synthesis of isopyochelin.
LC-MS analyses on a homochiral stationary phase of all the synthetic products and the
natural metabolites was performed using a ChiralPAK IE analytical column packed with
amylose tris (3,5-dichlorophenylcarbamate). Both synthetic products were injected
separately, followed by co-injection with the ethanol extract of the culture supernatant to
determine which included the stereoisomers corresponding to the metabolite (Figure 4.9).
Synthetic stereoisomers of isopyochelin with R configuration at C-4" can be completely
separated on the column. None of these had the same retention time as the natural
products. On the other hand, two of the synthetic diastereoisomers with S configuration
at C-4" had identical retention times to the natural products, although only three peaks
-53
-43
-33
-23
-13
-3
7
17
27
37
47
300280260240220200180
CD
(m
deg
)
Wavelength (nm)
2-Methyl-L-cysteine
2-Methyl-D-cysteine
Chapter 4 : Watasemycin biosynthesis
108
were observed for the synthetic compounds, indicating incomplete separation on the
column. Thus, the absolute configuration at C-4" in isopyochelin was established as S.
Figure 4.9. EICs at m/z = 325.0 (corresponding to [M+H]+ for isopyochelin) from
LC-HRMS analyses of the synthetic stereoisomers of isopyochelin and the natural
metabolites from the ethanol extract of S. coelicolor M1152/SV-2_E03::SspI culture
supernatant on a homochiral stationary phase. Comparison with 4"R-isopyochelins 135
(left) and 4"S-isopyochelins 136 (right) is shown separately.
4.3 Roles of Sven0508 and Sven0516 in the biosynthetic pathway
As discussed in Section 4.1.1, both sven0508 and sven0516 putatively encode PchK-like
thiazoline reductases in the sven0503-sven0517 gene cluster. However, it was proposed
that only one thiazoline reductase is necessary for the reduction step in watasemycin and
thiazostatin biosynthesis since they have the same scaffold structure as pyochelin and
enantio-pyochelin. Therefore, individual in-frame deletions of these genes were
constructed by Dr. Yuki Inahashi to investigate their roles in the production of 2-
hydroxyphenylthiazolines in S. venezuelae.
In order to see whether and how production of the 2-hydroxyphenylthiazoline metabolic
products would be affected by gene deletion of sven0508 or sven0516, the S. coelicolor
M1152/SV-2_E03::SspI and ∆sven0508 and ∆sven0516 mutant strains were grown on
both mannitol soya flour medium (SFM) and yeast-extract dextrose (YD) medium. Rich
medium SFM is commonly used as solid medium for spore production, with mannitol
being the primary carbon source. YD medium is a nutritionally complete medium, in
which yeast extract is the source of nitrogen and growth factors while dextrose provides
an energy source for the growth of microorganisms. YD medium was used as liquid
medium for heterologous expression of specialised metabolites in this study.
Chapter 4 : Watasemycin biosynthesis
109
4.3.1 Sven0516 is required for 2-hydroxyphenylthiazoline biosynthesis
The ethanol extract of the culture broth from the ∆sven0516 mutant was analysed by
LC-HRMS. None of the 2-hydroxyphenylthiazoline metabolic products could be
detected in the extract from the ∆sven0516 mutant (Figure 4.10). Thus it can be
concluded that the PchK homologue, Sven0516, is required for the biosynthesis of all the
2-hydroxyphenylthiazoline metabolic products of the sven0503-sven0517 gene cluster
and probably functions as a thiazoline reductase.
Figure 4.10. EICs at m/z = 353.0988, 339.0832, 210.0583, 224.0740 and 325.0675,
corresponding to [M+H]+ for watasemycin, thiazostatin, aerugine, pulicatins A/B and
isopyochelin, respectively, from LC-HRMS analyses of the ethanol extract of the
∆sven0516 mutant culture supernatant (purple traces), in comparison to the S. coelicolor
M1152/SV-2_E03::SspI strain (blue traces).
Interestingly, deletion of sven0516 also abolishes the production of aerugine 75 and
pulicatins A/B 77/78, the assembly of which ostensibly does not require a thiazoline
reductase. Thus we suspect that aerugine and pulicatins A/B are derived from hydrolytic
Chapter 4 : Watasemycin biosynthesis
110
cleavage and subsequent reduction of thiazostatin/isopyochelin and watasemycin,
respectively, rather than being direct products of the pathway (Scheme 4.8).
Scheme 4.8. Proposed formation of aerugine 75 and pulicatins A/B 77/78, from
hydrolysis and reduction of isopyochelin/thiazostatin and watasemycin, respectively.
4.3.2 Sven0508 is not involved in biosynthesis of 2-hydroxyphenylthiazolines
The ethanol extract of the culture broth from the ∆sven0508 mutant was also analysed
by LC-HRMS. Watasemycin 83, thiazostatin 82, aerugine 75, pulicatins A/B 77/78 and
isopyochelin were all still observed in the extract from the ∆sven0508 mutant, at
comparable levels to the wild type strain (Figure 4.11). This implied that the PchK
homologue, Sven0508, is not involved in the biosynthesis of the metabolic products of
the sven0503-sven0517 gene cluster.
Chapter 4 : Watasemycin biosynthesis
111
Figure 4.11. EICs at m/z = 353.0988, 339.0832, 210.0583, 224.0740 and 325.0675,
corresponding to [M+H]+ for watasemycin, thiazostatin, aerugine, pulicatins A/B and
isopyochelin, respectively, from LC-HRMS analyses of the ethanol extract of the
∆sven0508 mutant culture supernatant.
4.4 Reassignment of the absolute stereochemistry of watasemycin and thiazostatin
Watasemycin is structurally closely related to pyochelin and thiazostatin. The relative
stereochemistry of thiazostatin and watasemycin was assigned in the literature as shown
in Figure 4.12 on the basis of NOE NMR studies.117 As mentioned previously, PchK is
proposed to catalyse reduction of the second thiazoline, resulting in production of
enantio-pyochelin 76 which has S configuration at both C-4' and C-4''. Herein, we have
established that the absolute configuration at C-4" in isopyochelin is S. Since Sven0516
appears to function as the thiazoline reductase in the biosynthesis of all the
2-hydroxyphenylthiazolines from the sven0503-sven0517 gene cluster, including
thiazostatin and watasemycin, it is reasonable to speculate that they all have same S
absolute configuration at C-4". Combined with the previously assigned relative
stereochemistry for thiazostatin and watasemycin, the absolute configuration can be
deduced as 4'R, 5'R, 4"S for watasemycin and 4'R, 4"S for thiazostatin, as shown in the
structure circled in red in Figure 4.12.
Chapter 4 : Watasemycin biosynthesis
112
Figure 4.12. Two reported possible absolute configurations of watasemycin and
thiazostatin. The absolute stereochemistry with 4"S configuration is circled in red.
Besides thiazostatin and watasemycin, aerugine 75 and pulicatins A/B 77/78 are also
proposed to be the products of the sven0503-sven0517 gene cluster based on high
resolution LC-MS analyses. Their structures have been fully elucidated as discussed in
Section 1.5.3 (Figure 1.17). It was noticed that pulicatins A/B 77/78 have a methyl group
at C-5', which is the same as in watasemycin, and the absolute configuration at that
stereocenter has been confirmed to be S.115 This is contradictory to the assignment of the
absolute stereochemistry of watasemycin based on the relative stereochemistry reported
by Sasaki et al.117 Thus we suspect that their assignment of the relative stereochemistry
may be incorrect for watasemycin and thiazostatin. Moreover, NOE studies of
watasemycin showed that the methyl group at C-5' is proposed to be anti to the hydrogen
atom at C-4', suggesting their absolute configuration may be either 4'R, 5'R or 4'S, 5'S.
Taken together, the absolute stereochemistry can be reassigned as 4'S, 5'S, 4''S and 4'S,
4''S for watasemycin and thiazostatin, respectively, as shown in Figure 4.13.
Figure 4.13. Reassigned stereochemistry of thiazostatin, watasemycin and isopyochelin.
Chapter 4 : Watasemycin biosynthesis
113
An MT-like domain in PchE (or homologue) is believed to catalyse the epimerisation of
the stereocenter in the first thiazoline during pyochelin biosynthesis in P. aeruginosa and
S. scabies, resulting in the R configuration at C-4'.114,121 As discussed in Section 4.1.1,
bioinformatics analysis showed that this region in Sven0512 seems to be non-functional.
Lack of the MT-like domain in the PchE homologue in P. protegens leads to the
production of enantio-pyochelin 76 with 4'S configuration.114 The seemingly inactive
MT-like epimerisation domain in Sven0512 supports our assignment of a 4'S
configuration in watasemycin and thiazostatin. However, more work is needed to fully
elucidate their absolute stereochemistry.
4.5 Role of Sven0515 in watasemycin biosynthesis
4.5.1 Sven0515 methylates C-5' of thiazostatin to produce watasemycin
Bioinformatics analyses showed that the product of the sven0515 gene is homologous to
class B RS methylases, which are known to catalyse methylation on unreactive carbon
centres in the biosynthesis of a wide variety of specialised metabolites. Compared to
pyochelin, thiazostatin 82 contains one additional methyl group at C-4'' and besides this,
watasemycin 83 carrys another methyl group at C-5' (Figure 4.13). Based on its similarity
to class B RS methylases, Sven0515 is predicted to be responsible for introducing one or
both of these. In order to determine the specific role of Sven0515 in the biosynthesis of
2-hydroxyphenylthiazolines, a deletion mutant of sven0515 was also constructed by Dr.
Yuki Inahashi.
LC-MS analysis of the ethanol extract from the culture supernatant of the sven0515
mutant strain showed that it is able to produce aerugine 75, thiazostatin 82 and
isopyochelin 137, but not watasemycin 83 or pulicatins A/B 77/78, which all contain a
methyl group at C-5' (Figure 4.14). Thus it seems that Sven0515 is required for the
methylation of C-5', but not C-4'', due to the production of thiazostatin and isopyochelin,
both of which contain the methyl group at C-4''.
Chapter 4 : Watasemycin biosynthesis
114
Figure 4.14. EICs at m/z = 353.0988, 339.0832, 210.0583, 224.0740 and 325.0675,
corresponding to [M+H]+ for watasemycin, thiazostatin, aerugine, pulicatins A/B and
isopyochelin, respectively, from LC-MS analyses of the ethanol extract of the sven0515
culture supernatant (green traces), in comparison to the S. coelicolor
M1152/SV-2_E03::SspI strain (blue traces).
To investigate the timing of C-5' methylation, we fed the metabolites in the extract of the
sven0515 mutant to the sven0516 mutant. The sven0516 mutant was confirmed not
to produce any of the 2-hydroxyphenylthiazolines due to absence of the thiazoline
reductase as described in Section 4.3.2. Incubation of the culture for four days after
feeding with the extract of the sven0515 mutant resulted in nearly complete conversion
of thiazostatin 82 in the extract to watasemycin 83 (Figure 4.15), indicating that
thiazostatin is the substrate of Sven0515. It also suggests that Sven0515-catalysed
methylation of thiazostatin is the final step in watasemycin biosynthesis.
Chapter 4 : Watasemycin biosynthesis
115
Figure 4.15. EICs from LC-MS analyses of the ethanol extract from the ∆sven0516
mutant fed with the extract from the ∆sven0515 mutant.
4.5.2 Synthesis of stereospecifically deuterium-labelled cysteine
As discussed in Section 1.5.2, the biosynthesis of pyochelin and enantio-pyochelin
involves condensation of the salicylate with L-cysteine to afford the first thiazoline ring.
The biosynthesis of thiazostatin and watasemycin is predicted to be similar for this step.
The C-5' on the thiazoline ring of thiazostatin would then be methylated by Sven0515 as
the final step in watasemycin biosynthesis.
Based on the determination of 5'S configuration in watasemycin, there are two possible
stereochemical courses for this reaction (Scheme 4.9). Firstly, Sven0515 could
stereospecifically abstract the pro-S hydrogen atom from C-5', followed by appending of
a methyl group from MeCbl to the resulting carbon in a suprafacial process with retention
of stereochemistry, route A in Scheme 4.9. Alternatively, Sven0515 could
stereospecifically abstract the pro-R hydrogen atom from C-5', followed by appending a
methyl group from MeCbl to the resulting carbon in an antarafacial process with
inversion of stereochemistry, route B in Scheme 4.9. A third possibility is non-
stereospecific abstraction of either pro-S or pro-R hydrogen atom from C-5' resulting in
methylation via a mixture of routes A and B in Scheme 4.9. Base on reported studies
about other homologues of class B RS methylases, the methylation is predicted to
proceed with inversion of stereochemistry at the reacting carbon.
Chapter 4 : Watasemycin biosynthesis
116
Scheme 4.9. Possible stereochemical courses for Sven0515-catalysed methylation,
which could occur via primarily route A or route B, or via a mixture of routes A and B.
Fom3, one of the first recognised class B RS methylases, has been characterised in vitro
to catalyse the stereoselective methylation of 2-hydroxyethylphosphonate (HEP) to
afford S-2-hydroxypropylphosphonate (HPP),185-187 the penultimate step in the
biosynthesis of the broad spectrum antibiotic fosfomycin (Scheme 4.10).
Scheme 4.10. The stereoselective methylation catalysed by Fom3 in fosfomycin
biosynthesis.
Prior to its successful characterisation in vitro, in order to investigate the stereochemical
course during fosfomycin 138 biosynthesis, Hammerschmidt and co-workers
synthesised HEPs deuterium labelled at the C-2 position, [2,2-2H2]-HEP, (R)-[2-2H1]-
HEP and (S)-[2-2H1]-HEP, and fed them to the fosfomycin producing strain
Streptomyces fradiae (Figure 4.16).188-189 Deuterium incorporation into fosfomycin was
measured based by NMR spectroscopic analysis of (1R,2R)-(2-amino-l-
hydroxypropy1)phosphonic acid, which was produced from fosfomycin by reaction
with ammonia. When the two deuterium labelled compound [2,2-2H2]-HEP, or pro-S
deuterium-labelled compound (S)-[2-2H1]-HEP, were fed to S. fradiae, 34% and 32% of
the fosfomycin produced contained label, respectively. However, feeding with the pro-R
deuterium-labelled compound, (R)-[2-2H1]-HEP, resulted in no retention of the label.
Thus, a mechanism was proposed by van der Donk et al. that the pro-R hydrogen atom
Chapter 4 : Watasemycin biosynthesis
117
was removed from C-2 of HEP during the methylation catalysed by Fom3.187 This should
be followed by an antarafacial process of the resulting HEP radical reacting with MeCbl
to achieve the methyl group with S-configuration at the reacting carbon in HPP.
Figure 4.16. Feeding of stereospecifically deuterium-labelled substrates to determine the
stereochemical course of the C-methylation reaction in fosfomycin biosynthesis.
A second example of a methylation that occurs with stereochemical inversion is
biosynthesis of macrocyclic peptide antibiotic bottromycin, which harbors several non-
proteinogenic amino acids. The enzymes responsible for these reactions also belong to
the class B family of radical SAM methylases. In order to determine the mechanism of
the methylations, Kellenberger et al. carried out feeding experiments in Streptomyces
bottropensis with chemically synthesised isotope-labelled precursors.190 These
precursors included isotopomers of the corresponding amino acids, valine and
phenylalanine (Figure 4.17). The valine isotopomer carrys one methyl group labelled by
13C and the other by deuterium, while both of the phenylalanine isotopomers have a
stereo-defined deuterium label at the β position. The incorporation results show that the
methylations occur with inversion of configuration at the centres undergoing
functionalisation.
Chapter 4 : Watasemycin biosynthesis
118
Figure 4.17. Feeding of stereospecifically isotope-labelled amino acids to determine the
stereochemical course of the methylations in bottromycin biosynthesis, reported by
Kellenberger and Arigoni.190
The use of stereospecifically isotope-labelled substrates is an effective method for
elucidating the stereochemical course of enzymatic reactions. It was envisaged that a
similar approach could be employed to investigate the stereochemical course of the
methylation step catalysed by Sven0515 (Scheme 4.11).
Scheme 4.11. Proposed approach for mechanistic studies of Sven0515-mediated
methylation using stereospecifically deuterium-labelled L-cysteine.
Similar to pyochelin and enantio-pyochelin, two molecules of L-cysteine should be
incorporated into the structure of thiazostatin and watasemycin during biosynthesis. Thus,
feeding stereospecifically deuterium-labelled L-cysteine at the β-position to the
producing strain would lead to thiazostatin or watasemycin with a chiral label at the target
methylating position, C-5'and provide evidence to distinguish between the possible
scenarios in Scheme 4.9.
Chapter 4 : Watasemycin biosynthesis
119
Therefore, a synthetic method for the stereoselective incorporation of a deuterium atom
into the β-position of L-cysteine was required. Since cysteine is easily oxidised, it is
normally prepared as the homodimer form L-cystine, which is biochemically convertible
into its monomer L-cysteine. Examination of the literature revealed several methods for
the synthesis of cystine or cysteine stereospecifically deuterated or tritiated at the β-
position. However, most of the reported methods involve enzymatic and/or resolution
processes in order to obtain enantiomerically pure products.191-198 There are few
examples of purely chemical methods for introducing an isotope label in a stereospecific
manner.199-200 One reported method is reductive hydrogenation of the sterically hindered
and deuterated thiazoline with palladium followed by hydrolysis to achieve the desired
stereospecifically deuterium-labelled cysteine (Scheme 4.12).199 The modest total yield
of this synthetic approach, involving preparation of the thiazoline from corresponding
deuterium-labelled cysteine, made it too expensive and time-consuming for our purposes.
Scheme 4.12. Previous synthesis of stereospecifically deuterium-labelled L-cysteine at
the β-position through a deuterated thiazoline.199
Another approach was proposed by Maeda et al. to use a carbohydrate chirality template
with double steric constraint to achieve the chirally β-deuterated L-amino acids (Scheme
4.13).200
Scheme 4.13. Proposed method for synthesis of stereospecifically deuterium-labelled
L-cysteine at the β-position using a carbohydrate chirality template by Maeda et al.200
Chapter 4 : Watasemycin biosynthesis
120
More recently, Oba et al. developed another method for introducing a stereospecific
deuterium label at the β-position of L-cystine as outlined in Scheme 4.14.201 The
stereospecific deuterium label was obtained by transforming the carboxyl group of D-
serine into the stereospecifically deuterium-labelled alcohol via asymmetric reduction of
1-deuterio aldehyde. Most steps of this route were reported in excellent yields and the
reagents used can be easily handled. Due to the S-Alpine-Borane used in the reduction
step of the route, the stereochemistry of the deuterium label at the β-position was S-
configured and the diastereoselectivity of deuterium substitution was reported to be
approximately 95% de from integration of the 1H NMR spectrum. Based on these
advantages, we adopted this approach for our synthesis of the stereospecifically
deuterium-labelled L-cysteine (S-3-2H-L-cysteine 139) (Scheme 4.14).
Scheme 4.14. The synthetic route for (2R,2'R,3S,3'S)-[3,3'-2H2]cystine reported by Oba
et al.201
The 11-step synthesis begins with the preparation of [1-2H]serinal derivative 142. O-Bn
and N-Boc-protected D-serine, which is commercially available, was condensed with 3,5-
dimethylpyrazole and the pyrazolide 141 obtained was reduced with lithium aluminium
deuteride to give the required 1-deuterio aldehyde 142. To avoid racemisation of this N-
protected α-amino aldehyde,202 it was immediately used in the next step without
purification. This step requires careful quenching and control of the temperature to avoid
over-reduction. Asymmetric reduction of 1-deuterio aldehyde 142 using S-Alpine-
Chapter 4 : Watasemycin biosynthesis
121
Borane yielded stereospecifically deuterium-labelled alcohol 143 in more than 90% yield.
The mechanism of the reduction is described in Scheme 4.15. Diethanolamine was added
at the end of the reaction to precipitate the dialkyl borane byproduct.
Scheme 4.15. The mechanism of the stereospecific reduction of aldehyde 142 by
S-Alpine-Borane.
Subsequently, acetylation and debenzylation of the deuterated alcohol 143 followed by
RuO4-oxidation of the deprotected hydroxymethyl group of compound 145 afforded
protected L-serine derivative 146 in quantitative yield. The 1H NMR spectrum of 146, as
well as other doubly protected intermediates, had unusual signals and splitting patterns,
most notably for the protons at α- and β-position of the amino acid. This indicated the
presence of rotamers that are interchanging conformation on the timescale of the NMR
experiment. To investigate this hypothesis, the 1H NMR experiments for protected
L-serine derivative 146 were repeated at varying temperatures from 5 °C to 55 °C as
shown in Figure 4.18. As the temperature increased, conversion between the rotamers
became faster and the signal splitting was significantly reduced, confirming the presence
of rotamers.
Chapter 4 : Watasemycin biosynthesis
122
Figure 4.18. A comparison between 4.20-5.70 ppm of the 1H NMR spectra in CDCl3 at
600 MHz of (3R)-O-acetyl-N-(tert-butoxycarbonyl)-L-serine-3-2H 146 at varying
temperatures.
After protection of the carboxyl functionality as its tert-butyl ester, the deuterated L-
serine derivative 147 was transformed into the corresponding tosylate 149 in 55% yield
via deacetylation and tosylation. The tosylate 149 obtained was then treated with
potassium thioacetate to give deuterium-labelled L-cysteine derivative 150 in an SN2-
type displacement. Deprotection of compound 150 was carried out in refluxing 6 M HCl
solution and the resulting L-cysteine hydrochloride was purified by ion-exchange column
chromatography (Dowex® 50WX8 hydrogen form, 200-400 mesh). The eluent
containing L-cysteine was aerated in air for two days to give the (2R,2'R,3S,3'S)-[3,3'-
2H2]cystine 151 as a white solid.
The synthetic route proved to be reproducible and gave more than 20% total yield in our
hands. The 1H and 13C NMR spectra (Figure 4.19) of the product were consistent with
that reported by Oba et al.201
Chapter 4 : Watasemycin biosynthesis
123
Figure 4.19. The 1H (at 300 MHz) and 13C NMR (at 500 MHz) spectra of the synthesised
(2R,2'R,3S,3'S)-[3,3'-2H2]cystine 151 in 2.5% NaOD in D2O.
4.5.3 Incorporation of stereospecifically deuterium-labelled L-cystine into
watasemycin
With the successful synthesis of the stereospecifically deuterium-labelled L-cystine, the
incorporation approach to probe the stereochemical course of Sven0515-catalysed
methylation was investigated. S. coelicolor M1152/SV-2_E03::SspI was initially grown
in YD medium for two days to ensure good growth, then pelleted, transferred to
supplemented minimal medium (SMM), since this medium gives more equal production
levels of thiazostatin and watasemycin (Figure 4.20). SMM is a simple production
medium for secondary metabolites, with glucose as the primary carbon source.
Chapter 4 : Watasemycin biosynthesis
124
Figure 4.20. A comparison of EICs for thiazostatin (red traces) and watasemycin (black
traces) from the LC-MS analyses of the ethanol extracts of S. coelicolor
M1152/SV-2_E03::SspI culture supernatant grown in SMM and YD medium.
After incubation of the culture for 24 hours, labelled L-cystine was added. The feeding
procedure was repeated at 24 hour intervals until a total incubation time of 5 days at
30 °C. The synthesised labelled L-cystine fed into the culture could be easily converted
by bacteria to L-cysteine with the pro-S hydrogen atom replaced by a deuterium label at
the β-position.
Following extraction with ethanol, LC-MS analysis was carried out. Due to the
incorporation of two molecules of L-cysteine into thiazostatin and watasemycin, the
deuterium label may not only appear at the methylating position, C-5', but also C-5".
Therefore, feeding of deuterium-labelled cysteine should lead to production of
thiazostatin and its isotopomers harboring one deuterium label at C-5' or C-5", or two
deuterium labels at both of the positions (left column, Figure 4.21).
Chapter 4 : Watasemycin biosynthesis
125
Figure 4.21. Possible deuterium-labelled isotopomers of thiazostatin and watasemycin
produced after feeding with (2R, 3S)-[3-2H1] cysteine 139.
If Sven0515 stereospecifically abstracts the pro-R hydrogen atom from C-5', then feeding
of pro-S deuterium labelled L-cysteine should result in labelled watasemycin and its
isotopomers with retention of the pro-S deuterium label. As with thiazostatin, the
watasemycin isotopomers could include a deuterium label(s) at C-5' and/or C-5", as
shown in the middle column of Figure 4.21. The ratio of the watasemycin isotopomer
labelled with two deuterium to that labelled with one deuterium atoms should correspond
to the ratio of the thiazostatin isotopomers.
Alternatively, if Sven0515 stereospecifically abstracts the pro-S hydrogen atom from
C-5' of thiazostatin, then feeding of pro-S deuterium labelled cysteine should result in
loss of the deuterium at C-5' from watasemycin. In this case, there would not be a
watasemycin isotopomer labelled with two deuterium atoms but only isotopomers
labelled with one deuterium at C-5" (right column, Figure 4.21).
Finally, a non-stereospecific process would also result in production of isotopomer of
watasemycin with two deuterium labels. However, compared to the stereospecific
process, the ratio of the watasemycin isotopomer with two deuterium labels to the
isotopomer with one deuterium label should be lower than for thiazostatin, due to a
partial loss of the pro-S deuterium at C-5'.
Chapter 4 : Watasemycin biosynthesis
126
A comparison of the HRMS of thiazostatin and watasemycin resulting from LC-MS
analysis of the ethanol extracts of S. coelicolor M1152/SV-2_E03::SspI fed with
(2R,2'R,3S,3'S)-[3,3'-2H2]cystine 151 is shown in Figure 4.22. It shows that both
thiazostatin and watasemycin were produced as isotopomers harboring two deuterium
labels, which correspond to the peak at m/z = 341.0943 and m/z = 355.1099 for
thiazostatin and watasemycin, respectively.
Figure 4.22. A comparison of the HRMS for thiazostatin and watasemycin from LC-MS
analyses of the ethanol extracts of S. coelicolor M1152/SV-2_E03::SspI fed with
(2R,2'R,3S,3'S)-[3,3'-2H2]cystine.
In order to determine the relative amount of each isotopomer of thiazostatin and
watasemycin more accurately, a detailed analysis of the mass spectra is required. Taking
thiazostatin as an example (Figure 4.23), the peak at m/z = 340.0894 is a combination of
the natural isotopomer(s) of unlabelled thiazostatin and the singly deuterium-labelled
thiazostatin from our feeding experiment with deuterium-labelled cysteine. Natural
isotopomers of a compound exist due to the natural abundance of isotopes of the chemical
elements composing the molecule with the 13C-isotopomer mainly considered. The
calculation is necessary due to the inability of the mass spectrometer to distinguish
compounds containing a carbon-13 atom from those containing a deuterium atom.
Chapter 4 : Watasemycin biosynthesis
127
Figure 4.23. An example of deciphering the HRMS of the metabolic products from
feeding experiments with deuterium-labelled precursors, including the isotopic
distribution of natural thiazostatin (top) and singly deuterium-labelled thiazostatin
(middle) and the actual mass spectrum of thiazostatin detected (bottom).
The isotope distribution in each molecule we analysed was obtained from the LC-MS
data analysis software (Bruker Compass DataAnalysis). The relative amount of singly
deuterium-labelled thiazostatin (m/z = 340.0894) is determined as 49.9% of unlabelled
thiazostatin since 18.8% of the total peak at 340.0894 is due to the natural isotopomer(s)
of unlabelled thiazostatin (Table 4.2).
Table 4.2. Calculation of the relative amounts of the deuterium-labelled thiazostatins
from the HRMS.
m/z
[M+H]+
Relative amount
(%) of each peak
+1 Due to
natural isotopic
abundance
+2 Due to
natural isotopic
abundance
Relative amount (%)
of thiazostatins
339 100.0 - - 100 (0D)
340 68.7 100 × 18.8% - 49.9 (1D)
341 47.6 49.9 × 18.8% 100 × 11.2% 27.0 (2D)
For calculation of the relative amount of the doubly deuterium-labelled thiazostatin (m/z
= 341.0957 for [M+H]+ C15H17D2N2O3S2), it becomes more complicated due to
contribution of the +2 isotopomer(s) of unlabelled thiazostatin and the +1 isotopomer(s)
of singly deuterium-labelled thiazostatin. By subtracting these two contributions, the
amount of doubly deuterium-labelled thiazostatin was determined as 27.0% of unlabelled
Chapter 4 : Watasemycin biosynthesis
128
thiazostatin. Thus, the ratio of the doubly deuterium-labelled thiazostatin to singly
deuterium-labelled thiazostatin can be calculated as 0.54:1.
We then determined the relative amounts of singly deuterium- and doubly deuterium-
labelled watasemycins as 38.9% and 22.2%, respectively, compared to unlabelled
watasemycin (Table 4.3). The ratio of the doubly deuterium-labelled watasemycin to
singly deuterium-labelled watasemycin is 0.57:1.
Table 4.3. Calculation of the relative amounts of the deuterium-labelled watasemycins
from the HRMS.
m/z
[M+H]+
Relative amount
(%) of each peak
+1 due to
natural isotopic
abundance
+2 due to
natural isotopic
abundance
Relative amount (%)
of watasemycins
353 100.0 - - 100 (0D)
354 58.9 100 × 20.0% - 38.9 (1D)
355 39.3 38.9 × 20.0% 100 × 9.3% 22.2 (2D)
The fact that the ratio of doubly deuterium-labelled watasemycin to singly deuterium-
labelled watasemycin is very close to that of thiazostatin strongly suggests retention of
the pro-S deuterium label during Sven0515-catalysed methylation, as previously outlined
in Figure 4.21.
Although repeated feeding experiments produced varied levels of labelled thiazostatin
and watasemycin, the ratio in watasemycin was always comparable and often higher than
that in thiazostatin, further confirming the result. Thus it can be concluded that Sven0515
stereospecifically abstracts the pro-R hydrogen atom from C-5' and proceeds with
inversion of configuration at that position to give 5'S configuration (Figure 4.24).
Sven0515-catalysed methylation appears to share the same mechanism as other class B
RS methylases, in which inversion of stereochemistry also occurs. The result also
reciprocally supports our reassignment of the configuration at C-5' as S.
Chapter 4 : Watasemycin biosynthesis
129
Figure 4.24. Proposed stereochemical course during Sven0515-catalysed methylation at
C-5' of thiazostatin based on the feeding experiment.
In addition, it should be noted that the 3D peak (m/z 342 and 356 for [M+H]+ of
thiazostatin and watasemycin, respectively) in the HRMS of thiazostatin and
watasemycin appears more intense than expected. Examination of the L-cysteine
metabolic pathway revealed that the β-carbon is transferred to tetrahydrofolate to give
N5,N10-methylenetetrahydrofolate in primary metabolism, which is later incorporated
into homocysteine. Homocysteine is then used to regenerate methionine, which is a
necessary component for SAM. Thus, we suspect that part of the 3D peak may be due to
thiazostatin and watasemycin with two deuterium labels at C-5’ and C-5" and another
label at one of the methyl groups. Reducing the concentration of labelled cystine used in
feeding or co-feeding of L-methionine (2-4 mM) at the same time proved to be effective
for minimising the intensity of the 3D peak while retention of deuterium label at C-5'
was still observed with similar results as above.
4.6 Proposed pathway for watasemycin biosynthesis
Based on above results, a biosynthetic pathway for 2-hydroxyphenylthiazolines from the
sven0503-sven0517 gene cluster is proposed (Figure 4.25).
Chapter 4 : Watasemycin biosynthesis
130
Figure 4.25. Proposed pathway for the biosynthesis of thiazostatin 82, watasemycin 83,
isopyochelin 137, aerugine 75 and pulicatins A/B 77/78 in S. venezuelae ATCC 10712.
Sven0506 catalyses formation of salicylate from chorismate. Salicylate is adenylated by
the A domain, Sven0510, in the presence of ATP and loaded onto the N-terminal PCP
domain of the NRPS Sven0512. The internal A domain of Sven0512 catalyses
adenylation of the first molecule of L-cysteine, which is then covalently linked to its C-
terminal PCP domain. The Cy domain in Sven0512 catalyses condensation of the
thioesters of L-cysteine and salicylate and subsequent cyclisation to give the first
thiazoline ring. During this process, the stereochemistry of L-cysteine is retained due to
lack of a functional MT domain to carry out an epimerisation. A second molecule of
L-cysteine is adenylated by the A domain of Sven0517 and linked to its PCP domain.
The Cy domain of Sven0517 catalyses a similar condensation and cyclisation reaction to
generate a 2-hydroxyphenyl-bis-thiazolinyl thioester. Until this point, all the reactions
are analogous to the biosynthesis of enantio-pyochelin 76 in P. protegens (Figure 1.21).
The MT domain in Sven0517 probably catalyses the methylation at C-4" with inversion
of stereochemistry. The thiazoline reductase Sven0516 then catalyses reduction of the
Chapter 4 : Watasemycin biosynthesis
131
second thiazoline ring. The selective reduction of the second thiazoline by Sven0516
seems to be due to the opposite stereochemistry of the two cysteine-derived stereocenters.
Thioester hydrolysis catalysed by the C-terminal TE domain of Sven0517 at this stage
results in the production of isopyochelin. Hydrolysis after N-methylation by the MT
domain of Sven0517 leads to the production of thiazostatin, which is methylated by the
class B RS methylase Sven0515 to yield watasemycin. Aerugine 75 is proposed to be
produced from hydrolytic cleavage and reduction of thiazostatin/isopyochelin while
pulicatins A/B 77/78 derive from watasemycin in the same way.
Chapter 4 : Watasemycin biosynthesis
132
4.7 Summary
Bioinformatics analyses of the S. venezuelae ATCC10712 genome sequence revealed the
presence of a putative biosynthetic gene cluster, sven0503-sven0517, for pyochelin-like
metabolites. Heterologous expression of this gene cluster in S. coelicolor M1152 resulted
in identification and characterisation of the main metabolic products, watasemycin and
thiazostatin, as well as other 2-hydroxyphenylthiazolines, including aerugine and
pulicatins A/B, which are putatively identified by high resolution LC-MS. A novel
compound, isopyochelin, has also been found to be one of the metabolic products from
the gene cluster. The absolute configuration of isopyochelin at C-4" has been
unambiguously established as S based on comparison with synthetic standards.
Both Sven0508 and Sven0516 share sequence similarities to the thiazoline reductase
PchK. Gene deletion of sven0508 did not affect the production of all the
2-hydroxyphenylthiazolines from the sven0503-sven0517 gene cluster, suggesting that it
is not involved in the biosynthetic pathway. Deletion of sven0516 resulted in abolishment
of all the 2-hydroxyphenylthiazoline metabolic products from the gene cluster, indicating
it plays an essential role as the putative thiazoline reductase. A deletion mutant of
sven0515, which putatively encodes a homologue of class B RS methylases, lost the
ability to produce watasemycin and pulicatins A/B, implying a critical role of Sven0515
in methylation at C-5'. Further feeding experiments, supplying the extract from the
sven0515 strain to the sven0516 strain showed complete conversion of thiazostatin to
watasemycin, suggesting that thiazostatin is the substrate for Sven0515-catalysed
methylation.
The absolute stereochemistry of watasemycin and thiazostatin has been reassigned as 4'S,
5'S, 4"S and 4'S, 4"S, respectively, based on the experimental results and analysis. The
reassignment relies on the assumption that watasemycin and thiazostatin have the same
4"S configuration as isopyochelin. Moreover, based on these results, a biosynthetic
pathway has been proposed for the 2-hydroxyphenylthiazolines from the sven0503-
sven0517 gene cluster in S. venezuelae ATCC10712.
The stereochemical course of the methylation catalysed by Sven0515 on the chemically
inert 5' carbon of thiazostatin has been investigated by incorporation experiments using
chemically synthesised L-cysteine stereospecifically deuterium labelled at the β-position.
LC-MS analyses showed that the pro-R hydrogen atom at C-5' of thiazostatin is
Chapter 4 : Watasemycin biosynthesis
133
substituted by the transferred methyl group while the pro-S deuterium is retained. The
stereochemical course of the Sven0515-catalysed methylation reaction therefore
proceeds with inversion of configuration, consistent with the mechanism of C-
methylation by other class B RS methylases. This also reciprocally supports the
reassignment of the stereochemistry of watasemycin 83 and thiazostatin 82.
Chapter 5 : Conclusions and future work
134
5. Conclusions and future work
Chapter 5 : Conclusions and future work
135
5.1 MmfL catalyses the formation of a butenolide intermediate in AHFCA
biosynthesis
A TBDMS-protected derivative of the MmfL reaction product has been synthesised
using the methodology reported by Sello and co-workers,168 involving condensation of
an appropriate acyl Meldrum’s acid and TBDMS mono-protected dihydroxyacetone to
give the corresponding β-ketoester, which cyclises during purification by flash
chromatography on silica. Deprotection using a hydrogen fluoride-pyridine complex led
to the desired butenolide, which was confirmed by its ESI-MS/MS analysis. It should be
noted that the TBDMS-protected butenolide is unstable and degraded quickly to a species
corresponding to the deprotected butenolide.
The dephosphorylated product of the MmfL-catalysed reaction of DHAP and the NAC
β-ketothioester has been identified to be the butenolide by comparing to the synthetic
standard. Thus, it can be concluded that MmfL catalyses the reaction between a β-
ketothioester and DHAP to form a phosphorylated butenolide intermediate, which is an
analogous reaction to that catalysed by AfsA during the γ-butyrolactone signalling
molecule A-factor biosynthesis in S. griseus.67 Therefore, the formation of GBLs and
AHFCAs are proposed to proceed via a shared butenolide phosphate intermediate
(Scheme 5.1).74-75 Each type of enzyme (i.e., AfsA, MmfL or ScbA) presumably has a
different preference for the length of the alkyl chain in the β-ketothioester substrate.
Scheme 5.1. General scheme for the biosynthesis of Streptomyces signalling molecules,
GBLs and AHFCAs, which are proposed to share a butenolide phosphate intermediate.
As the biosynthetic roles of both MmfL and MmfP have been established in vitro, it
would be of particular interest to study the function of another key enzyme, MmfH, in
AHFCA biosynthesis. Based on previous work,77 one promising way would be to
elucidate the X-ray crystal structure of MmfH.
Chapter 5 : Conclusions and future work
136
5.2 Interactions between DNA, ligand and ArpA-like repressors
5.2.1 Three operators for MmfR binding
Three conserved MAREs within the methylenomycin biosynthetic gene cluster have
been confirmed as the binding sites for MmfR, including the operator upstream of mmyR
and the operators located in the mmyY-mmyB, mmfR-mmfL intergenic regions. The order
of affinity of MmfR for the three operators has been shown in vitro as follows: the
operator upstream of mmyB > the operator between mmfR and mmfL > the operator
upstream of mmyR.
This order partially supports the proposed regulatory cascade in methylenomycin
biosynthesis. When AHFCA production reaches a critical concentration, they bind to the
MmfR repressor and cause derepression of MmfR binding to the operator within the
mmfL-mmfR intergenic region, leading to a significant increase of the production of
AHFCAs. The increased concentration of AHFCAs is then able to release MmfR from
the mmyB promoter, which has the highest affinity, and thus the transcriptional activator
MmyB can be produced. MmyB binds to its operator regions (‘B-boxes’), initiating the
transcription of the structural genes to produce methylenomycins.
Previous gene deletion experiments showed that MmyR-like repressors likely act as an
off-switch to stop the production of specialised metabolites, and appear to be
unresponsive to signalling molecules. Due to lack of direct evidence on the function of
MmyR, further studies are required to fully reveal the mechanism of the transcriptional
regulation involving MmfR/MmyR/AHFCAs in methylenomycin biosynthesis.
5.2.2 Molecular basis for AHFCA recognition by MmfR
Site-directed mutagenesis based on the X-ray crystal structure of MmfR with AHFCA2
bound was used to investigate ligand binding and recognition by MmfR. Tyr85Ala,
Tyr85Phe, Tyr144Ala, Tyr144Phe and Gln130Glu mutants of MmfR were prepared and
their interaction with DNA and AHFCA1 were assessed in vitro using EMSAs. Key
residues involved in interactions with AHFCAs in MmfR were identified, in particular
tyrosines at positions 85 and 144, glutamine at position 130 and leucine at position 110.
Chapter 5 : Conclusions and future work
137
The hydroxyl group of Tyr85 and the amide group in the side chain of Gln130 were
confirmed to play important roles in the interaction of MmfR with the ligand through
hydrogen bond formation with the carboxyl group of AHFCAs. Alteration of these
residues attenuates the ability of the ligand to release MmfR from its operators. For
Tyr144, it is likely that its backbone amide participates in ligand recognition and binding
while its side chain may also be involved in the formation of the ligand binding pocket.
Leu110 is probably important for proper folding and stability of MmfR. These results
have illuminated the molecular basis for AHFCA recognition by MmfR.
5.2.3 MmfR is not responsive to γ-butyrolactones
A representative of the γ-butyrolactone class of signalling molecules, SCB1, was
synthesised in this study to determine whether there is any cross-talk between the
regulation systems mediated by GBLs and AHFCAs. The effect of SCB1 on the binding
of MmfR to its operator was investigated in vitro using EMSAs. The data showed that
SCB1 is unable to release MmfR from its operator. This demonstrated that ligand binding
for MmfR is highly specific, supporting the hypothesis that GBLs and AHFCAs mediate
pathway-specific regulation of antibiotic biosynthesis in S. coelicolor A3(2).
5.2.4 SgnR in the regulation of gaburedin biosynthesis
As a close homologue of MmfR, SgnR was proposed to be able to bind AHFCA ligands.
In silico analysis also identified three ARE sequences in the gaburedin biosynthetic gene
cluster, which were proposed to be the binding sites for SgnR. These predictions have
been experimentally confirmed by in vitro studies using EMSAs.
It has been shown that SgnR is able to bind the MARE sequence from S. coelicolor in a
pre-test, and the three ARE sequences from the S. venezuelae gbn gene cluster, including
the operator upstream of gbnA and the operators located in the sgnR-sgnL, gbnR-sgnH
intergenic regions. The SgnR-DNA complexes can be dissociated from its operators by
addition of AHFCA6, indicating that gaburedin biosynthesis in S. venezuelae is regulated
by a similar system to that used for the regulation of methylenomycin biosynthesis in S.
coelicolor.
Chapter 5 : Conclusions and future work
138
Additionally, the key amino acids in MmfR that have been demonstrated in the above
studies to be important for ligand binding or protein stability, are conserved in SgnR
(Figure 5.1), further supporting the similar roles played by SgnR and MmfR in regulation
of specialised metabolite biosynthesis.
Figure 5.1. Sequence alignment of SgnR and MmfR. The key amino acid residues
identified from MmfR are highlighted in yellow.
5.3 Watasemycin biosynthesis in S. venezuelae ATCC10712
5.3.1 Isopyochelin is a shunt metabolite in the watasemycin biosynthetic pathway
Bioinformatics analyses of the S. venezuelae ATCC 10712 genome sequence revealed
the presence of a putative biosynthetic gene cluster, sven0503-sven0517, for pyochelin-
like metabolites. Most of the genes encode proteins with high sequence similarity to those
encoded by genes within the S. scabies pyochelin biosynthetic gene cluster. Heterologous
expression of the gene cluster in S. coelicolor M1152 revealed thiazostatin and
watasemycin as the main metabolic products of the cluster, as well as other putative
2-hydroxyphenylthiazolines identified by LC-MS, including aerugine and pulicatins A/B.
A metabolite with molecular formula identical to pyochelin was also identified.
Comparison to synthetic standards showed that it is actually not pyochelin, but
isopyochelin, which is a structural isomer of pyochelin with the C-4" position methylated
instead of the nitrogen atom of the thiazolidine. The configuration at C-4" of
isopyochelin was established as S by comparing the metabolite to synthetic stereoisomers
by chiral LC-MS.
Chapter 5 : Conclusions and future work
139
5.3.2 Sven0516 is the only thiazoline reductase required for watasemycin
biosynthesis
Sven0516 is similar in sequence to PchK from P. protegens, which has been proposed to
function as a thiazoline reductase in enantio-pyochelin biosynthesis. Deletion of
sven0516 (carried out by Dr. Yuki Inahashi) abolished the production of all the
2-hydroxyphenylthiazoline metabolic products, indicating its key role as a thiazoline
reductase in the watasemycin biosynthetic pathway. Production of aerugine and
pulicatins, which ostensibly do not require a thiazoline reductase for their biosynthesis,
was also abolished in the mutant strain, suggesting that they may be derived from
hydrolysis and reduction from thiazostatin/isopyochelin and watasemycin, respectively.
The presence of a PchK-like thiazoline reductase is consistent with the assignment of S
configuration at C-4"of isopyochelin, as enantiopyochelin has the same S configuration
at C-4''.
Another PchK homologue, Sven0508, has been found not to be involved in watasemycin
biosynthesis since deletion of the gene encoding it resulted in a similar metabolic profile
to the wild type strain (results of Dr. Yuki Inahashi).
5.3.3 Sven0515 catalyses C-methylation of thiazostatin to produce watasemycin
Deletion of sven0515 (carried out by Dr. Yuki Inahashi) abolished the production of
pulicatins and watasemycin, both of which contain a methyl group at C-5' on the
thiazoline ring, with no effect on the production of thiazostatin, aerugine and
isopyochelin. This revealed an essential role for Sven0515, a class B RS methylase, in
the C-methylation of the thiazoline ring in watasemycin biosynthesis. Feeding of natural
thiazostatin extracted from the culture of the ∆sven0515 mutant to the ∆sven0516 strain
resulted in its complete transformation to watasemycin, indicating that thiazostatin is the
substrate of Sven0515. The Sven0515-catalysed methylation of thiazostatin is the final
step in watasemycin biosynthesis.
Class B RS methylases are known to catalyse methylation of un-activated carbon centres
in the biosynthesis of a wide variety of natural products, including aminoglycosides,
aminocoumarins, β-lactams, enediynes, polyketides, phosphonates, and RiPPs. To our
knowledge, the Sven0515-catalysed methylation of thiazostatin is the first example of
Chapter 5 : Conclusions and future work
140
such a reaction that has been experimentally validated to be found in the biosynthesis of
a nonribosomal peptide.
The stereochemical course of the methylation on the chemically inert 5' carbon of
thiazostatin has been investigated by incorporation experiments using chemically
synthesised L-cystine with the pro-S hydrogen atom replaced by a deuterium atom at the
β-position. LC-MS analyses showed retention of the deuterium label at C-5' in
watasemycin, indicating that the pro-R hydrogen atom of thiazostatin is substituted by
the appended methyl group while the pro-S deuterium is retained. The stereochemical
inversion during the Sven0515-catalysed methylation is consistent with other members
of class B RS methylases, such as GenD1 in gentamicin biosynthesis and MoeK5 in
moenomycin biosynthesis.149,165
Feeding with synthetic L-cysteine with the pro-R hydrogen atom replaced by a deuterium
atom at the β-position would further support the proposed stereochemical course. In this
case, the pro-R deuterium atom should be abstracted during methylation, resulting in
singly deuterium-labelled watasemycin. The synthesis of this deuterium-labelled
cysteine could be achieved by using the same synthetic approach with the corresponding
stereospecific reducing agent, R-Alpine-Borane (Scheme 5.2).201
Scheme 5.2. Synthetic route to L-cysteine with the pro-S or pro-R hydrogen atom
replaced by a deuterium atom at the β-position.
Enzymatic assays in vitro using the purified and reconstituted Sven0515 enzyme and
synthetic thiazostatin substrate harboring a stereospecifically-incorporated deuterium
label at the site of methylation would provide direct evidence for the stereochemical
course of the methylation reaction. Furthermore, despite the important role played by
class B RS methylases in antibiotic biosynthesis, no X-ray crystal structures of enzymes
belonging to this family have been reported to date. Thus, it will be useful to determine
Chapter 5 : Conclusions and future work
141
the X-ray crystal of Sven0515, a member of this family of methylases, which would
provide a clear picture of the structure-function relationships of these enzymes. Due to
the oxygen-sensitive nature of class B RS methylases, crystallisation trials for Sven0515
would need to be set up in an anaerobic chamber.
This would develop a better understanding of the catalytic mechanism employed by class
B RS methylases to methylate inert carbon centres. Such reactions, which are key steps
in the biosynthesis of numerous natural products, are currently impossible using
anthropogenic methodologies. Thus, the information obtained will underpin
development of new methodologies in biotechnology and production of new or
improved antimicrobial compounds.
5.3.4 Stereochemistry of thiazostatin and watasemycin
The relative stereochemistry of thiazostatin and watasemycin has been reported before
this study. However, their absolute stereochemistry hasn’t been elucidated since their
discovery. Our absolute stereochemical assignment of thiazostatin and watasemycin is
inferred on the basis of the 4"S absolute configuration determined for isopyochelin, as
they are all derived from the same biosynthetic pathway.
Thus, it will be important to experimentally confirm the absolute configuration at C-4"
of thiazostatin, which is the substrate for Sven0515. This could be achieved in the same
way as the configuration of isopyochelin was determined, involving synthesis of
thiazostatin stereoisomers, followed by subsequent chiral LC-MS comparison with the
natural metabolites (Scheme 5.3).
Scheme 5.3. Proposed route to characterise the absolute configuration at C-4" of
thiazostatin.
Chapter 5 : Conclusions and future work
142
The configuration at C-5' of watasemycin could be characterised using the same
methodology. However, condensation with a β-methyl-cysteine, instead of a normal
cysteine, in the first step is required to introduce a methyl group at C-5' of the final
product (Scheme 5.4).
Scheme 5.4. Proposed route to characterise the absolute configuration at C-5' of
watasemycin.
During the synthesis of pyochelin, it has been noticed that epimerisation occurs at the
C-4' position under the reaction conditions, which has been reported in the literature.181-
182 This makes preparation of an authentic standard with a specific configuration at C-4'
impossible. Moreover, it appears that feeding of isotope-labelled precursors, such as
2-2H-L-cysteine, would also be ineffective, due to the fact that the deuterium atom at the
α-position of the cysteine may be washed out by primary metabolism before
incorporation into the biosynthesis, e.g., by a pyridoxal phosphate (PLP)-mediated
exchange reaction.203 However, the C-4' configuration can be deduced from the finding
that the C-4' hydrogen substituent and the C-5' methyl group in watasemycin are anti to
each other on the basis of NOE studies.
Chapter 6 : Experimental
143
6. Experimental
Chapter 6 : Experimental
144
6.1 Synthetic chemistry
6.1.1 General information
Dry solvents, including THF, DMF, DCM, Et2O and toluene, were obtained from the
solvent towers in the O’Reilly Group (Department of Chemistry, University of Warwick)
and stored over activated 4 Å molecular sieves under argon. All other solvents were
purchased from Sigma-Aldrich, Fisher Scientific or VWR and used without further
purification. All reagents were purchased from Sigma-Aldrich, Fisher Scientific, VWR,
Alfa Aesar, Acros Organic, Scientific Laboratory Supplies, Atlantic Research Chemicals
or Carbosynth and used without further purification. All air- and moisture-sensitive
manipulations were carried out with standard Schlenk techniques under argon.
Flash column chromatography was conducted on Sigma-Aldrich silica gel (40-63 μm,
60 Å). TLC was performed on aluminium backed sheets pre-coated with Merck silica gel
60 F254 and visualised by UV radiation, potassium permanganate or ninhydrin stains.
Preparative TLC was performed on Analtech 20 cm × 20 cm silica gel plate with a
thickness of 2000 μm on glass support for purification of small amounts of product which
are UV-visible. Solvents were evaporated using a Buchi Rotavapor R-200 equipped with
a Buchi Vacuum pump V-700.
The 1H- and 13C-NMR spectra were obtained using a Bruker AV-250 MHz, Avance III
HD 300, 400 or 500 MHz spectrometer. The variable temperature NMR experiments
were obtained using a Bruker AV-600 MHz spectrometer. Deuterated solvents were
purchased from Sigma-Aldrich and used as supplied. All chemical shifts are given as δ
values in ppm with reference to residual HDO (δH 4.79), CHCl3 (δH 7.26) or TMS (δH
0.00) if residue CHCl3 peak was difficult to be distinguished (e.g. overlapping with other
peaks), the central peak of CDCl3 (δC 77.00) or DMSO-d6 (δC 39.52). Data for 1H NMR
spectra are reported as follows: chemical shift (δ, ppm), multiplicity (s = singlet, d =
doublet, t = triplet, dd = doublet of doublets, td = triplet of doublets, ddd = doublet of
doublet of doublets, br = broad, m = multiplet), integration, assignment and coupling
constant (Hz).
Low resolution ESI mass spectra were recorded using an Agilent 6130B single quad
spectrometer. HRMS spectra were measured by the University of Warwick mass
Chapter 6 : Experimental
145
spectrometry service on Bruker MaXis or MaXis impact mass spectrometer. Melting
points were measured in open capillaries using a Stuart SMP10 melting point apparatus
and were uncorrected.
General procedure for condensation to yield pyochelin and isopyochelin
To a solution of 2-(2-hydroxyphenyl)-4,5-dihydrothiazole-4-carbaldehyde (25 mg,
0.12 mmol) in a mixture of 5 mL ethanol and 1.3 mL water was added CH3COOK
(82 mg, 0.84 mmol) and the hydrochloride salt of the corresponding methylated cysteine
(30 mg, 0.17 mmol). After being stirred at RT for 24 hours, 15 mL water was added and
the mixture washed once with hexane. The aqueous phase was then acidified to pH 5 and
extracted with EtOAc. The combined EtOAc layers were washed with brine, dried over
MgSO4 and concentrated under vacuum to yield crude pyochelin or isopyochelin as a
yellow solid. Due to the formation of four diastereoisomers in this step and the fact that
the free carboxylic acids cannot be isolated in pure form due to their rapid
interconversion, NMR assignments of the product were difficult. Therefore, these
compounds were analysed by LC-MS directly.
General procedure for acyl Meldrum’s acid synthesis starting from carboxylic acid
To a solution of carboxylic acid (1 eq) in DCM was added EDC.HCl (1.5 eq) and the
mixture cooled to 0 °C. Then DMAP (3 eq) dissolved in DCM was added, followed by
Meldrum’s acid (1 eq) dissolved in DCM. After stirring at RT for 18 hours, the reaction
mixture was diluted with DCM, washed with 1 M HCl, water and brine. The organic
phase was dried over MgSO4 and concentrated under vacuum. The residue was purified
by column chromatography to yield the product.
General procedure for butenolide synthesis
A solution of monosilyl DHA (1 eq) and acyl Meldrumʼs acid (1.2 eq) in 2 mL toluene
was refluxed for 3 hours at which point was added more acyl Meldrumʼs acid (0.5 eq).
After stirring for an additional 5 hours, the reaction mixture was allowed to stand
at -20 °C for 19 hours and purified by column chromatography (hexanes/acetone = 60:1
v/v) without removal of toluene to yield the desired product.
Chapter 6 : Experimental
146
General procedure for butenolide reduction
To a solution of NaBH3CN (2 eq) in acetic acid was dropwise added a solution of
butenolide (1 eq) in acetic acid and the mixture stirred at RT for 8 hours. After removal
of acetic acid under vacuum, the oily residue was diluted with EtOAc and washed with
5% NaHCO3 solution and brine. The organic phase was then dried over MgSO4 and
concentrated under vacuum. The residue was purified by flash column chromatography
to yield the product.
The reduction would result in the formation of three stereogenic centres in the molecule
so NMR spectra contain peaks from a mixture of different stereoisomers. This results in
excessive peaks and inaccurate integration. The diastereoisomers were isolated in the
next step after removal of the silyl group.
General procedure for silyl deprotection of γ-butyrolactones
A 0.1 M solution of TBDMS-protected γ-butyrolactone in a 6:3:1 mixture of
THF/HCOOH/H2O was stirred at RT overnight. The mixture was adjusted to pH 4 with
saturated NaHCO3 solution and extracted with EtOAc. The combined EtOAc layers were
dried over MgSO4 and concentrated under vacuum. The diastereoisomers were purified
and separated by column chromatography (pet ether/EtOAc = 1:1 v/v).
General procedure for β-ketoester synthesis
A solution of acyl Meldrum’s acid in methanol was refluxed overnight. After removal of
the solvent under vacuum, the residue was purified by column chromatography to afford
the desired β-ketoester.
General procedure for AHFCA cyclisation
To a 0.5 M solution of β-ketoester (1 eq) and 2,2-dimethyl-1,3-dioxan-5-one (1 eq) in
methanol was added scandium III triflate (0.1 eq) and the reaction mixture stirred at RT
overnight. After removal of the solvent under vacuum, the residue was purified by
column chromatography to give the desired furan.
Chapter 6 : Experimental
147
General procedure for hydrolysis of AHFCA methyl ester
To a solution of the furan ester (1 eq) in a 1:1 mixture of THF/water was added LiOH
(2.5 eq) and the reaction mixture stirred at RT overnight. After removal of the THF under
vacuum, the remaining aqueous phase was washed with hexane, acidified with 1 M HCl
until a white precipitate formed, and extracted with Et2O. The combined organic layers
were washed with brine, dried over MgSO4 and concentrated to afford the product.
General procedure for β-ketothioester synthesis
To a 0.5 M solution of acyl Meldrum’s acid (1.5 eq) in 1,2-dichloroethane under argon
was added N-acetylcysteamine (1 eq) and the reaction mixture refluxed for 3 hours. After
removal of the solvent under vacuum, the residue was purified by column
chromatography with a plug of silica impregnated with CuSO4 (10 cm) on top to afford
the desired product.
6.1.2 Synthesis of monosilyl DHA
1-((tert-Butyldimethylsilyl)oxy)-3-hydroxypropan-2-one 102
To a mixture of dihydroxyacetone (5.00 g, 55.6 mmol) and imidazole (1.51 g,
22.2 mmol) in 50 mL DMF at 0 °C was added a solution of tert-butyldimethylsilyl
chloride (2.67 g, 17.8 mmol) in 10 mL DMF and the reaction stirred at RT for 17 hours.
Then 50 mL water was added and the mixture extracted with Et2O. The organic phase
was washed with brine, dried over MgSO4 and concentrated under vacuum. The residue
was purified by column chromatography (pet ether/EtOAc = 5:1 v/v) to give 1-((tert-
butyldimethylsilyl)oxy)-3-hydroxypropan-2-one (1.82 g, 50%) as a colourless oil.
1H NMR (300 MHz, CDCl3) δ: 0.12 (s, 6H, H-4), 0.90 (s, 9H, H-6), 2.99 (br s, 1H, O-H),
4.31 (s, 2H, H-1), 4.50 (s, 2H, H-3), further signals at 3.43 - 4.24 due to formation of the
intermolecular ketal compounds.
13C NMR (75 MHz, CDCl3) δ: 210.88 (C-2), 25.74 (C-6), 18.15 (C-5), -5.56 (C-4),
further signals at 63.07 - 111.38.
Chapter 6 : Experimental
148
HR-MS: m/z calculated for C9H20NaO3Si [M+Na]+ : 227.1074; found: 227.1076.
6.1.3 Synthetic MmfL substrate and enzymatic product
6.1.3.1 Synthesis of NAC β-ketothioester
2,2-Dimethyl-5-(3-methylbutanoyl)-1,3-dioxane-4,6-dione 99
To a solution of Meldrum’s acid (1.79 g, 12.4 mmol) in 25 mL DCM was added
triethylamine (2.52 g, 25.0 mmol) and DMAP (0.30 g, 2.5 mmol) and the mixture stirred
for 10 mins at RT before being cooled to 0 °C. Then 3-methylbutanoyl chloride (1.50 g,
12.4 mmol) was added dropwise and the reaction warmed to RT and stirred overnight.
The resulting mixture was washed with 1 M HCl and water, dried over MgSO4 and
concentrated under vacuum. The residue was purified by column chromatography (pet
ether/EtOAc = 2:1 v/v) to yield 2,2-dimethyl-5-(3-methylbutanoyl)-1,3-dioxane-4,6-
dione (2.80 g, 99%) as a yellow oil.
1H NMR (300 MHz, CDCl3) δ: 0.89 (d, 6H, H-1, J = 7.0 Hz), 1.61 (s, 6H, H-8), 2.03 -
2.12 (m, 1H, H-2), 2.85 (d, 2H, H-3, J = 7.0 Hz), 15.19 (br s, 1H, O-H)
13C NMR (75 MHz, CDCl3) δ: 170.22 (C-4), 159.89 (C-6), 104.33 (C-7), 91.62 (C-5),
43.45 (C-3), 27.05 (C-2), 26.40 (C-8), 22.17 (C-1)
LR-MS: m/z calculated for C11H15O5 [M-H]- found: 227.1.
S-(2-Acetamidoethyl) 5-methyl-3-oxohexanethioate 61
Following the general procedure for β-ketothioester synthesis, work up gave S-(2-
acetamidoethyl) 5-methyl-3-oxohexanethioate (0.21 g, 60%) as a yellowish oil. The
keto-enol form ratio was nearly 2:1, estimated from integration of the H-5 and H-5' peaks
Chapter 6 : Experimental
149
in 1H NMR.
1H NMR (400 MHz, CDCl3) δ: 0.89 (d, 6H, H-1, J = 6.5 Hz), 1.94 (s, 3H, H-10), 1.99
(br s, 2H, H-3'), 2.11 (nonet, 1H, H-2, J = 6.5 Hz), 2.37 (d, 2H, H-3, J = 7.0 Hz), 3.05 (t,
2H, H-7, J = 6.0 Hz), 3.42 (t, 2H, H-8, J = 6.0 Hz), 3.65 (s, 2H, H-5), 5.40 (s, 1H, H-5'),
6.23 (br s, 1H, N-H), 12.55 (br s, 1H, O-H)
13C NMR (100 MHz, CDCl3) δ: 194.08 (C-4'), 192.21 (C-4), 176.54 (C-6), 170.51 (C-9),
170.36 (C-9'), 100.04 (C-5'), 57.45 (C-5), 52.10 (C-3), 43.95 (C-3'), 39.74 (C-8'), 39.04
(C-8), 29.04 (C-7), 27.67 (C-7'), 26.33 (C-2), 24.23 (C-10), 22.29 (C-1)
HR-MS: m/z calculated for C11H19NNaO3S [M+Na]+: 268.0978; found: 268.0985.
5-(1-Hydroxyhexylidene)-2,2-dimethyl-1,3-dioxane-4,6-dione 100
Following the general procedure for acyl Meldrum’s acid synthesis, work up gave 5-(1-
hydroxyhexylidene)-2,2-dimethyl-1,3-dioxane-4,6-dione (1.67 g, 67%) as a yellow oil.
1H NMR (400 MHz, CDCl3) δ: 0.78 (t, 3H, H-1, J = 7.0 Hz), 1.17 - 1.31 (m, 4H, H-2 and
3), 1.54 - 1.58 (m, 2H, H-4), 1.61(s, 6H, H-10), 2.93 (t, 2H, H-5, J = 7.5 Hz), 15.18 (br
s, 1H, O-H)
13C NMR (100 MHz, CDCl3) δ: 170.23 (C-6), 159.77 (C-8), 104.37 (C-9), 90.96 (C-7),
35.28 (C-5), 31.14 (C-3), 26.38 (C-10), 25.46 (C-4), 21.93 (C-2), 13.48 (C-1)
LR-MS: m/z calculated for C12H17O5 [M-H]- found: 241.1.
S-(2-Acetamidoethyl) 3-oxooctanethioate 101
Following the general procedure for β-ketothioester synthesis, work up gave S-(2-
acetamidoethyl) 3-oxooctanethioate (0.35 g, 80%) as a white solid. The keto-enol form
Chapter 6 : Experimental
150
ratio was nearly 3:1, estimated from integration of the H-5 and H-5' peaks in 1H NMR.
1H NMR (400 MHz, CDCl3) δ: 0.86 - 0.90 (m, 3H, H-1), 1.24 - 1.31 (m, 4H, H-2 and 3),
1.58 (quintet, 2H, H-4, J = 7.0 Hz), 1.96 (s, 3H, H-12), 2.16 (t, 2H, H-5', J = 7.5 Hz),
2.51 (t, 2H, H-5, J = 7.5 Hz), 3.05 - 3.09 (m, 2H, H-9), 3.42 - 3.48 (m, 2H, H-10), 3.68
(s, 2H, H-7), 5.45 (s, 1H, H-7'), 6.02 (br s, 1H, N-H), 12.59 (br s, 1H, O-H)
13C NMR (100 MHz, CDCl3) δ: 194.24 (C-6'), 192.34 (C-6), 177.63 (C-8), 170.41 (C-11),
99.09 (C-7'), 57.14 (C-7), 43.37 (C-5), 39.86 (C-5'), 39.14 (C-10), 34.82 (C-10'), 31.19
(C-9'), 31.07 (C-9), 29.17 (C-3), 27.78 (C-3'), 25.86 (C-4'), 23.11 (C-12), 23.05 (C-4),
22.32 (C-2), 22.28 (C-2'), 13.82 (C-1)
HR-MS: m/z calculated for C12H21NNaO3S [M+Na]+: 282.1134; found: 282.1147.
Melting point: 89-90 °C.
6.1.3.2 Synthetic enzymatic product
4-(((tert-Butyldimethylsilyl)oxy)methyl)-3-hexanoylfuran-2(5H)-one 106
Following the general procedure for butenolide synthesis, work up gave 4-(((tert-
butyldimethylsilyl)oxy)methyl)-3-hexanoylfuran-2(5H)-one (0.17 g, 30%) as a
yellowish oil.
1H NMR (400 MHz, CDCl3) δ: 0.08 (s, 6H, H-6), 0.81 - 0.92 (m, 12H, H-8 and 14), 1.29
- 1.32 (m, 4H, H-12 and 13), 1.59 (quintet, 2H, H-11, J = 7.0 Hz), 2.95 (t, 2H, H-10, J =
7.5 Hz), 4.97 (s, 2H, H-5), 5.05 (s, 2H, H-4)
13C NMR (100 MHz, CDCl3) δ: 197.08 (C-9), 181.16 (C-1), 170.61 (C-3), 122.43 (C-2),
70.17 (C-4), 61.81 (C-5), 41.52 (C-10), 31.19 (C-11), 30.80 (C-7), 25.65 (C-8), 22.80
(C-12), 22.44 (C-13), 13.85 (C-14), -5.70 (C-6)
HR-MS: m/z calculated for C17H30NaO4Si [M+Na]+: 349.1806; found: 349.1801.
Chapter 6 : Experimental
151
3-Hexanoyl-4-(hydroxymethyl)furan-2(5H)-one 107
To a solution of 4-(((tert-butyldimethylsilyl)oxy)methyl)-3-hexanoylfuran-2(5H)-one
(2 mg) in 200 μL THF at 0 °C was added 2 drops of hydrogen fluoride-pyridine and the
mixture stirred for 40 mins at 0 °C. The reaction was quenched with 500 μL saturated
NaHCO3 solution, concentrated under vacuum to remove the THF and extracted with
EtOAc. The organic layers were combined, concentrated and redissolved in 700 μL of
1:1 v/v methanol/water. The sample was then filtered (0.4 μm) for LC-MS analyses.
HR-MS: m/z calculated for C11H16NaO4 [M+Na]+: 235.0941; found: 235.0948.
6.1.4 Synthesis of SCB6
5-(1-Hydroxyheptylidene)-2,2-dimethyl-1,3-dioxane-4,6-dione 115
Following the general procedure for acyl Meldrum’s acid synthesis, work up gave 5-(1-
hydroxyheptylidene)-2,2-dimethyl-1,3-dioxane-4,6-dione (1.13 g, 55%) as a yellow oil.
1H NMR (300 MHz, CDCl3) δ: 0.76 (t, 3H, H-1, J = 6.5 Hz), 1.13 - 1.24 (m, 6H, H-2, 3
and 4), 1.29 (quintet, 2H, H-5, J = 7.0 Hz), 1.61 (s, 6H, H-11), 2.94 (t, 2H, H-6, J =
7.5 Hz), 15.19 (br s, 1H, O-H)
13C NMR (75 MHz, CDCl3) δ: 170.25 (C-7), 159.79 (C-9), 104.38 (C-10), 90.95 (C-8),
35.36 (C-6), 31.07 (C-3), 28.71 (C-4), 26.39 (C-11), 25.77 (C-5), 22.12 (C-2), 13.65 (C-
1)
HR-MS: m/z calculated for C13H21O5 [M+H]+: 257.1384; found: 257.1374.
Chapter 6 : Experimental
152
4-(((tert-Butyldimethylsilyl)oxy)methyl)-3-heptanoylfuran-2(5H)-one 116
Following the general procedure for butenolide synthesis, work up gave 4-(((tert-
butyldimethylsilyl)oxy)methyl)-3-heptanoylfuran-2(5H)-one (125 mg, 30%) as a yellow
oil.
1H NMR (300 MHz, CDCl3) δ: 0.10 (s, 6H, H-6), 0.88 (t, 3H, H-15, J = 5.0 Hz), 0.90 (s,
9H, H-8), 1.25 - 1.29 (m, 6H, H-12, 13 and 14), 1.57 - 1.66 (m, 2H, H-11), 2.97 (t, 2H,
H-10, J = 7.5 Hz), 4.99 (s, 2H, H-5), 5.06 (s, 2H, H-4)
13C NMR (100 MHz, CDCl3) δ: 70.20 (C-4), 61.84 (C-5), 41.60 (C-10), 31.59 (C-13),
28.72 (C-12), 25.66 (C-8), 23.07 (C-11), 22.46 (C-14), 13.99 (C-15), C-1, 2, 3, 6, 7 and
9 signals missing
HR-MS: m/z calculated for C18H33O4Si [M+H]+: 341.2143; found: 341.2125.
4-(((tert-Butyldimethylsilyl)oxy)methyl)-3-(1-hydroxyheptyl)dihydrofuran-2(3H)-
one 117
Following the general procedure for butenolide synthesis, work up gave 4-(((tert-
butyldimethylsilyl)oxy)methyl)-3-(1-hydroxyheptyl)dihydrofuran-2(3H)-one (77 mg,
61%) as a colourless oil.
HR-MS: m/z calculated for C18H36NaO4Si [M+Na]+: 367.2275; found: 367.2281.
SCB6 stereoisomers 118
Following the general silyl deprotection procedure, work up and column chromatography
gave a pair of SCB6 and its enantiomer (24 mg, 47%, higher polarity) and a pair of their
Chapter 6 : Experimental
153
diastereoisomers (20 mg, 39%, lower polarity) as colourless oils.
(3R,4R)-3-((R)-1-Hydroxyheptyl)-4-(hydroxymethyl)dihydrofuran-2(3H)-one,
SCB6
(3S,4S)-3-((S)-1-Hydroxyheptyl)-4-(hydroxymethyl)dihydrofuran-2(3H)-one,
SCB6 enantiomer
1H NMR (400 MHz, CDCl3) δ: 0.88 (t, 3H, H-13, J = 7.0 Hz), 1.25 - 1.34 (m, 8H, H-9,
10, 11 and 12), 1.47 - 1.63 (m, 2H, H-8), 2.49 (br s, 2H, H-6 and 14), 2.65 (dd, 1H, H-2,
J = 9.5 and 4.5 Hz), 2.72 - 2.82 (m, 1H, H-3), 3.68 (dd, 1H, H-5a, J = 10.5 and 6.5 Hz),
3.75 (dd, 1H, H-5b, J = 10.5 and 5.0 Hz), 3.98 (t, 1H, H-4a, J = 9.0 Hz), 4.02 (ddd, 1H,
H-7, J = 9.0, 4.5 and 3.0 Hz), 4.42 (t, 1H, H-4b, J = 8.5 Hz)
13C NMR (125 MHz, CDCl3) δ: 177.21 (C-1), 70.86 (C-7), 68.33 (C-4), 63.00 (C-5),
49.05 (C-2), 40.10 (C-3), 33.95 (C-8), 31.70 (C-11), 29.08 (C-9), 25.76 (C-10), 22.56
(C-12), 14.04 (C-13)
HR-MS: m/z calculated for C12H22NaO4 [M+Na]+: 253.1410; found: 253.1411.
SCB6 diastereoisomers
(3R,4R)-3-((S)-1-Hydroxyheptyl)-4-(hydroxymethyl)dihydrofuran-2(3H)-one
(3S,4S)-3-((R)-1-Hydroxyheptyl)-4-(hydroxymethyl)dihydrofuran-2(3H)-one
1H NMR (400 MHz, CDCl3) δ: 0.88 (t, 3H, H-13, J = 7.0 Hz), 1.25 - 1.35 (m, 8H, H-9,
10, 11 and 12), 1.45 - 1.64 (m, 2H, H-8), 2.00 (br s, 2H, H-6 and 14), 2.56 (dd, 1H, H-2,
J = 7.5 and 3.5 Hz), 2.81 - 2.90 (m, 1H, H-3), 3.71 (dd, 1H, H-5a, J = 10.5 and 6.0 Hz),
Chapter 6 : Experimental
154
3.75 (dd, 1H, H-5b, J = 10.5 and 5.5 Hz), 4.08 - 4.14 (m, 2H, H-4a and 7), 4.42 (t, 1H,
H-4b, J = 8.5 Hz)
13C NMR (125 MHz, CDCl3) δ: 178.41 (C-1), 70.80 (C-7), 69.37 (C-4), 63.32 (C-5),
48.05 (C-2), 38.07 (C-3), 34.81 (C-8), 31.71 (C-11), 29.06 (C-9), 25.76 (C-10), 22.56
(C-12), 14.03 (C-13)
HR-MS: m/z calculated for C12H22NaO4 [M+Na]+: 253.1410; found: 253.1411.
6.1.5 Synthesis of 6-methylheptanoic acid
Dimethyl 2-(4-methylpentyl)malonate 119
To a suspension of NaH (0.72 g, 30.0 mmol) in 60 mL THF at 0 °C was slowly added
dimethyl malonate (2.90 g, 22.0 mmol) and the mixture stirred at the same temperature
for 30 mins before 1-bromo-4-methylpentane (3.30 g, 20.0 mmol) and TBAI (1.48 g,
4.0 mmol) were added. After being stirred at 0 °C for an additional 30 mins, the reaction
was allowed to warm to RT and heated to reflux overnight. The mixture was concentrated,
acidified with 3 M HCl and extracted with DCM. The combined organic layers were
washed with water and brine, dried over MgSO4 and concentrated under vacuum. The
residue was purified by column chromatography (pet ether/EtOAc = 10:1 v/v) to give
dimethyl 2-(4-methylpentyl)malonate (3.54 g, 82%) as a colourless oil.
1H NMR (300 MHz, CDCl3) δ: 0.80 (d, 6H, H-1, J = 6.5 Hz), 1.09 - 1.17 (m, 2H, H-3),
1.20 - 1.30 (m, 2H, H-4), 1.48 (nonet, 1H, H-2, J = 6.5 Hz), 1.82 (quartet, 2H, H-5, J =
8.0 Hz), 3.31 (t, 1H, H-6, J = 7.5 Hz), 3.68 (s, 6H, H-8)
13C NMR (75 MHz, CDCl3) δ: 169.81 (C-7), 52.25 (C-8), 51.55 (C-6), 38.25 (C-3),
28.88 (C-5), 27.52 (C-2), 24.98 (C-4), 22.33 (C-1)
HR-MS: m/z calculated for C11H21O4 [M+H]+: 217.1434; found: 217.1441.
Chapter 6 : Experimental
155
2-(4-Methylpentyl)malonic acid 120
To a solution of KOH (6.40 g, 114.3 mmol) in a mixture of 20 mL water and 40 mL
methanol was added dimethyl 2-(4-methylpentyl)malonate (1.35 g, 6.3 mmol) and the
resulting mixture stirred at RT overnight. After removal of the methanol under vacuum,
the remaining aqueous solution was washed with Et2O, acidified with 3 M HCl and
extracted with DCM. The combined organic layers were dried over MgSO4 and
concentrated to yield 2-(4-methylpentyl)malonic acid (1.17 g, 100%) as white crystals.
1H NMR (300 MHz, CDCl3) δ: 0.87 (d, 6H, H-1, J = 6.5 Hz), 1.17 - 1.25 (m, 2H, H-3),
1.34 - 1.44 (m, 2H, H-4), 1.55 (nonet, 1H, H-2, J = 6.5 Hz), 1.92 (quartet, 2H, H-5, J =
8.0 Hz), 3.44 (t, 1H, H-6, J = 7.5 Hz), 10.73 (br s, 2H, COO-H)
LR-MS: m/z calculated for C9H15O4 [M-H]- found: 187.1.
6-Methylheptanoic acid 121
A solution of 2-(4-methylpentyl)malonic acid (2.93 g, 15.6 mmol) in 30 mL DMF was
refluxed overnight. Then 30 mL of 5% NaHCO3 solution was added and the mixture
washed with hexane, acidified with 3 M HCl and extracted with DCM. The combined
DCM layers were washed with brine, dried over MgSO4 and concentrated under vacuum
to give 6-methylheptanoic acid (2.20 g, 98%) as a yellowish oil.
1H NMR (300 MHz, CDCl3) δ: 0.85 (d, 6H, H-1, J = 7.0 Hz), 1.14 - 1.21 (m, 2H, H-3),
1.27 - 1.38 (m, 2H, H-4), 1.52 (nonet, 1H, H-2, J = 6.5 Hz), 1.60 (quintet, 2H, H-5, J =
7.5 Hz), 2.34 (t, 2H, H-6, J = 7.5 Hz), 11.73 (br s, 1H, COO-H)
13C NMR (75 MHz, CDCl3) δ: 180.68 (C-7), 38.49 (C-3), 34.14 (C-6), 27.76 (C-2), 26.81
(C-4), 24.86 (C-5), 22.49 (C-1)
HR-MS: m/z calculated for C8H17O2 [M+H]+: 145.1223; found: 145.1222.
Chapter 6 : Experimental
156
6.1.6 Synthesis of SCB1
5-(1-Hydroxy-6-methylheptylidene)-2,2-dimethyl-1,3-dioxane-4,6-dione 122
Following the general procedure for acyl Meldrum’s acid synthesis, work up gave 5-(1-
hydroxy-6-methylheptylidene)-2,2-dimethyl-1,3-dioxane-4,6-dione (2.05 g, 48%) as a
yellow oil.
1H NMR (300 MHz, CDCl3) δ: 0.79 (d, 6H, H-1, J = 6.5 Hz), 1.12 - 1.17 (m, 2H, H-3),
1.28 - 1.38 (m, 2H, H-4), 1.47 (nonet, 1H, H-2, J = 6.5 Hz), 1.61 (quintet, 2H, H-5, J =
7.5 Hz), 1.66 (s, 6H, H-11), 2.99 (t, 2H, H-6, J = 7.5 Hz), 15.23 (br s, 1H, O-H)
13C NMR (75 MHz, CDCl3) δ: 170.25 (C-7), 159.79 (C-9), 104.38 (C-10), 90.96 (C-8),
38.16 (C-3), 35.42 (C-6), 27.49 (C-2), 26.86 (C-4), 26.42 (C-11), 26.06 (C-5), 22.23 (C-
1)
LR-MS: m/z calculated for C14H21O5 [M-H]- found: 269.2.
4-(((tert-Butyldimethylsilyl)oxy)methyl)-3-(6-methylheptanoyl)furan-2(5H)-one
123
Following the general procedure for butenolide synthesis, work up gave 4-(((tert-
butyldimethylsilyl)oxy)methyl)-3-(6-methylheptanoyl)furan-2(5H)-one (172 mg, 33%)
as a yellow oil.
1H NMR (300 MHz, CDCl3) δ: 0.09 (s, 6H, H-6), 0.84 (d, 6H, H-15, J = 6.5 Hz), 0.89 (s,
9H, H-8), 1.10 - 1.36 (m, 4H, H-12 and 13), 1.47 - 1.63 (m, 3H, H-11 and 14), 2.96 (t,
2H, H-10, J = 7.5 Hz), 4.98 (s, 2H, H-5), 5.06 (s, 2H, H-4)
Chapter 6 : Experimental
157
13C NMR (75 MHz, CDCl3) δ: 197.08 (C-9), 181.23 (C-1), 170.64 (C-3), 122.41 (C-2),
70.19 (C-4), 61.83 (C-5), 41.64 (C-10), 38.69 (C-13), 29.66 (C-7), 27.81 (C-14), 26.85
(C-12), 25.65 (C-8), 23.35 (C-11), 22.55 (C-15), -5.68 (C-6)
HR-MS: m/z calculated for C19H35O4Si [M+H]+: 355.2299; found: 355.2306.
4-(((tert-Butyldimethylsilyl)oxy)methyl)-3-(1-hydroxy-6-methylheptyl)
dihydrofuran-2(3H)-one 124
Following the general procedure for butenolide reduction, work up gave 4-(((tert-
butyldimethylsilyl)oxy)methyl)-3-(1-hydroxy-6-methylheptyl)dihydrofuran-2(3H)-one
(85 mg, 49%) as a colourless oil.
HR-MS: m/z calculated for C19H38NaO4Si [M+Na]+: 381.2432; found: 381.2439.
SCB1 stereoisomers 125
Following the general silyl deprotection procedure, work up and column chromatography
gave a pair of SCB1 and its enantiomer (25 mg, 43%, higher polarity) and a pair of their
diastereoisomers (17 mg, 29%, lower polarity) as colourless oils.
(3R,4R)-3-((R)-1-Hydroxy-6-methylheptyl)-4-(hydroxymethyl)dihydrofuran-
2(3H)-one, SCB1
(3S,4S)-3-((S)-1-Hydroxy-6-methylheptyl)-4-(hydroxymethyl)dihydrofuran-2(3H)-
one, SCB1 enantiomer
1H NMR (400 MHz, CDCl3) δ: 0.86 (d, 6H, H-13, J = 6.5 Hz), 1.14 - 1.20 (m, 2H, H-11),
1.26 - 1.39 (m, 4H, H-9 and 10), 1.47 - 1.55 (m, 2H, H-8), 1.57 - 1.64 (m, 1H, H-12),
Chapter 6 : Experimental
158
2.65 (dd, 1H, H-2, J = 9.5 and 4.5 Hz), 2.74 - 2.81 (m, 1H, H-3), 3.68 (dd, 1H, H-5a, J =
10.5 and 6.5 Hz), 3.75 (dd, 1H, H-5b, J = 10.5 and 5.0 Hz), 3.98 (t, 1H, H-4a, J = 9.0 Hz),
4.01 (ddd, 1H, H-7, J = 9.0, 4.5 and 3.0 Hz), 4.42 (t, 1H, H-4b, J = 8.5 Hz), O-H signal
missing
13C NMR (100 MHz, CDCl3) δ: 70.82 (C-7), 68.32 (C-4), 62.98 (C-5), 49.14 (C-2), 40.08
(C-3), 38.82 (C-11), 33.94 (C-8), 27.88 (C-12), 27.18 (C-10), 26.08 (C-9), 22.60 (C-13),
C-1 signal missing
HR-MS: m/z calculated for C13H24NaO4 [M+Na]+: 267.1567; found: 267.1565.
SCB1 diastereoisomers
(3R,4R)-3-((S)-1-Hydroxy-6-methylheptyl)-4-(hydroxymethyl)dihydrofuran-
2(3H)-one
(3S,4S)-3-((R)-1-Hydroxy-6-methylheptyl)-4-(hydroxymethyl)dihydrofuran-2(3H)-
one
1H NMR (300 MHz, CDCl3) δ: 0.86 (d, 6H, H-13, J = 6.5 Hz), 1.14 - 1.21 (m, 2H, H-11),
1.28 - 1.38 (m, 4H, H-9 and 10), 1.47 - 1.62 (m, 3H, H-8 and 12), 1.98 (br s, 2H, H-6 and
14), 2.57 (dd, 1H, H-2, J = 7.5 and 4.0 Hz), 2.80 - 2.92 (m, 1H, H-3), 3.71 (dd, 1H, H-5a,
J = 10.0 and 4.5 Hz), 3.75 (dd, 1H, H-5b, J = 10.0 and 5.0 Hz), 4.08 - 4.15 (m, 2H, H-4a
and 7), 4.42 (t, 1H, H-4b, J = 8.5 Hz)
13C NMR (75 MHz, CDCl3) δ: 178.50 (C-1), 70.80 (C-7), 69.42 (C-4), 63.32 (C-5), 48.10
(C-2), 38.85 (C-11), 38.09 (C-3), 34.85 (C-8), 27.89 (C-12), 27.17 (C-10), 26.07 (C-9),
22.59 (C-13)
HR-MS: m/z calculated for C13H24NaO4 [M+Na]+: 267.1567; found: 267.1565.
Chapter 6 : Experimental
159
6.1.7 Synthesis of AHFCAs
6.1.7.1 AHFCA1
Methyl 5-methyl-3-oxohexanoate 110
Following the general procedure for β-ketoester synthesis, work up gave methyl
5-methyl-3-oxohexanoate (1.59 g, 82%) as a yellow oil.
1H NMR (400 MHz, CDCl3) δ: 0.78 (d, 6H, H-1, J = 6.5 Hz), 1.97 - 2.04 (m, 1H, H-2),
2.28 (d, 2H, H-3, J = 7.0 Hz), 3.30 (s, 2H, H-5), 3.58 (S, 3H, H-7)
13C NMR (100 MHz, CDCl3) δ: 202.08 (C-4), 167.33 (C-6), 51.84 (C-7), 51.44 (C-3),
49.01 (C-5), 23.94 (C-2), 22.02 (C-1)
HR-MS: m/z calculated for C8H14NaO3 [M+Na]+: 181.0835, found: 181.0831.
Methyl 4-(hydroxymethyl)-2-isobutylfuran-3-carboxylate 113
Following the general procedure for AHFCA cyclisation, work up gave methyl
4-(hydroxymethyl)-2-isobutylfuran-3-carboxylate (0.17 g, 63%) as a colourless oil.
1H NMR (400 MHz, CDCl3) δ: 0.90 (d, 6H, H-10, J = 6.5 Hz), 2.02 (nonet, 1H, H-9, J =
7.0 Hz), 2.79 (d, 2H, H-8, J = 7.0 Hz), 3.68 (br s, 1H, O-H), 3.85 (s, 3H, H-7), 4.53 (s,
2H, H-1), 7.23 (s, 1H, H-3)
13C NMR (75 MHz, CDCl3) δ: 165.73 (4), 164.17 (C-6), 138.42 (C-3), 125.71 (C-2),
112.69 (C-5), 55.74 (C-1), 51.64 (C-7), 36.93 (C-8), 28.31 (C-9), 22.37 (C-10)
HR-MS: m/z calculated for C11H16NaO4 [M+Na]+: 235.0941, found: 235.0941.
Chapter 6 : Experimental
160
4-(Hydroxymethyl)-2-isobutylfuran-3-carboxylic acid, AHFCA1 26
Following the general procedure for AHFCA methyl ester hydrolysis, work up gave
4-(hydroxymethyl)-2-isobutylfuran-3-carboxylic acid (0.11 g, 70%) as a yellowish solid.
1H NMR (400 MHz, CDCl3) δ: 0.93 (d, 6H, H-9, J = 6.5 Hz), 2.09 (nonet, 1H, H-8, J =
7.0 Hz), 2.88 (d, 2H, H-7, J = 7.0 Hz), 4.61 (s, 2H, H-1), 7.28 (s, 1H, H-3), 7.80 (br s,
2H, O-H and COO-H)
13C NMR (100 MHz, CDCl3) δ: 170.15 (C-4), 166.17 (C-6), 138.79 (C-3), 125.27 (C-2),
112.15 (C-5), 55.58 (C-1), 36.79 (C-7), 28.25 (C-8), 22.35 (C-9)
HR-MS: m/z calculated for C10H14NaO4 [M+Na]+: 221.0784, found: 221.0786.
Melting point: 91-93 °C.
6.1.7.2 AHFCA3
5-(1-Hydroxy-4-methylpentylidene)-2,2-dimethyl-1,3-dioxane-4,6-dione 111
Following the general procedure for acyl Meldrum’s acid synthesis, work up gave 5-(1-
hydroxy-4-methylpentylidene)-2,2-dimethyl-1,3-dioxane-4,6-dione (1.05 g, 84%) as a
yellow oil.
1H NMR (300 MHz, CDCl3) δ: 0.91 (d, 6H, H-1, J = 6.0 Hz), 1.48 - 1.58 (m, 2H, H-3),
1.64 (nonet, 1H, H-2, J = 6.5 Hz), 1.70 (s, 6H, H-9), 3.03 (t, 2H, H-4, J = 7.5 Hz), 15.26
(br s, 1H, O-H)
13C NMR (75 MHz, CDCl3) δ: 198.54 (C-5), 170.49 (C-7), 104.66 (C-8), 91.07 (C-6),
34.83 (C-3), 33.83 (C-4), 28.04 (C-2), 26.69 (C-9), 22.11 (C-1)
Chapter 6 : Experimental
161
LR-MS: m/z calculated for C12H17O5 [M-H]- found: 241.2.
Methyl 6-methyl-3-oxoheptanoate 112
Following the general procedure for β-ketoester synthesis, work up gave methyl
6-methyl-3-oxoheptanoate (0.71 g, 95%) as a yellow oil.
1H NMR (400 MHz, CDCl3) δ: 0.77 (d, 6H, H-1, J = 6.5 Hz), 1.35 - 1.39 (m, 2H, H-3),
1.44 (nonet, 1H, H-2, J = 6.5 Hz), 2.43 (t, 2H, H-4, J = 7.5 Hz), 3.36 (s, 2H, H-6), 3.61
(s, 3H, H-8)
13C NMR (100 MHz, CDCl3) δ: 202.69 (C-5), 167.44 (C-7), 51.93 (C-8), 48.64 (C-6),
40.77 (C-4), 31.88 (C-3), 27.21 (C-2), 21.99 (C-1)
HR-MS: m/z calculated for C9H16NaO3 [M+Na]+: 195.0992, found: 195.0998.
Methyl 4-(hydroxymethyl)-2-isopentylfuran-3-carboxylate 114
Following the general procedure for AHFCA cyclisation, work up gave methyl
4-(hydroxymethyl)-2-isopentylfuran-3-carboxylate (0.19 g, 57%) as a yellow oil.
1H NMR (400 MHz, CDCl3) δ: 0.89 (d, 6H, H-11, J = 5.5 Hz), 1.46 - 1.54 (m, 3H, H-9
and 10), 2.90 (t, 2H, H-8, J = 7.5 Hz), 3.70 (br s, 1H, O-H), 3.84 (s, 3H, H-7), 4.51 (s,
2H, H-1), 7.20 (s, 1H, H-3)
13C NMR (75 MHz, CDCl3) δ: 165.68 (C-4), 165.10 (C-6), 138.25 (C-3), 125.78 (C-2),
111.87 (C-5), 55.71 (C-1), 51.67 (C-7), 36.92 (C-8), 27.67 (C-10), 26.22 (C-9), 22.23
(C-11)
HR-MS: m/z calculated for C12H18NaO4 [M+Na]+: 249.1097, found: 249.1095.
Chapter 6 : Experimental
162
4-(Hydroxymethyl)-2-isopentylfuran-3-carboxylic acid, AHFCA3 42
Following the general procedure for AHFCA methyl ester hydrolysis, work up gave
4-(hydroxymethyl)-2-isopentylfuran-3-carboxylic acid (0.11 g, 65%) as a yellowish
solid.
1H NMR (400 MHz, CDCl3) δ: 0.85 (d, 6H, H-10, J = 5.5 Hz), 1.47 - 1.50 (m, 3H, H-8
and 9), 2.92 (t, 2H, H-7, J = 7.0 Hz), 4.53 (s, 2H, H-1), 7.20 (s, 1H, H-3), 7.55 (br s, 2H,
O-H and COO-H)
13C NMR (100 MHz, CDCl3) δ: 169.85 (C-4), 167.00 (C-6), 138.60 (C-3), 125.24 (C-2),
111.36 (C-5), 55.51 (C-1), 36.67 (C-7), 27.68 (C-9), 26.19 (C-8), 22.22 (C-10)
HR-MS: m/z calculated for C11H16NaO4 [M+Na]+: 235.0941, found: 235.0948.
Melting point: 75-76 °C.
6.1.7.3 AHFCA6
5-(1-Hydroxy-5-methylhexylidene)-2,2-dimethyl-1,3-dioxane-4,6-dione 126
Following the general procedure for acyl Meldrum’s acid synthesis, work up gave 5-(1-
hydroxy-5-methylhexylidene)-2,2-dimethyl-1,3-dioxane-4,6-dione (1.35 g, 53%) as a
yellow oil.
1H NMR (250 MHz, CDCl3) δ: 0.87 (d, 6H, H-1, J = 6.5 Hz), 1.23 - 1.32 (m, 2H, H-3),
1.57 (nonet, 1H, H-2, J = 6.5 Hz), 1.62 - 1.69 (m, 2H, H-4), 1.71 (s, 6H, H-10), 3.03 (t,
2H, H-5, J = 7.5 Hz), 15.28 (br s, 1H, O-H)
Chapter 6 : Experimental
163
13C NMR (75 MHz, CDCl3) δ: 170.51 (C-6), 160.11 (C-8), 104.68 (C-9), 91.18 (C-7),
38.40 (C-3), 35.81 (C-5), 27.64 (C-2), 26.72 (C-10), 23.96 (C-4), 22.36 (C-1)
HR-MS: m/z calculated for C13H20NaO5 [M+Na]+: 279.1203, found: 279.1217.
Methyl 7-methyl-3-oxooctanoate 127
Following the general procedure for β-ketoester synthesis, work up gave methyl
7-methyl-3-oxooctanoate (0.93 g, 95%) as a yellow oil.
1H NMR (300 MHz, CDCl3) δ: 0.73 (d, 6H, H-1, J = 6.5 Hz), 0.99 - 1.13 (m, 2H, H-3),
1.33 - 1.50 (m, 3H, H-2 and 4), 2.39 (t, 2H, H-5, J = 7.5 Hz), 3.32 (s, 2H, H-7), 3.58 (s,
3H, H-9)
13C NMR (75 MHz, CDCl3) δ: 202.41 (C-6), 167.34 (C-8), 51.81 (C-9), 48.59 (C-7),
42.82 (C-5), 37.85 (C-3), 27.47 (C-2), 22.07 (C-1), 20.95 (C-4)
HR-MS: m/z calculated for C10H18NaO3 [M+Na]+: 209.1148, found: 209.1149.
Methyl 4-(hydroxymethyl)-2-(4-methylpentyl)furan-3-carboxylate 129
Following the general procedure for AHFCA cyclisation, work up gave methyl
4-(hydroxymethyl)-2-(4-methylpentyl)furan-3-carboxylate (0.24 g, 85%) as a colourless
oil.
1H NMR (300 MHz, CDCl3) δ: 0.85 (d, 6H, H-12, J = 6.5 Hz), 1.15 - 1.26 (m, 2H, H-10),
1.48 - 1.68 (m, 3H, H-9 and 11), 2.89 (t, 2H, H-8, J = 7.5 Hz), 3.57 (br s, 1H, O-H), 3.85
(s, 3H, H-7), 4.54 (s, 2H, H-1), 7.22 (s, 1H, H-3)
13C NMR (75 MHz, CDCl3) δ: 165.66 (C-4), 164.85 (C-6), 138.25 (C-3), 125.73 (C-2),
111.93 (C-5), 55.69 (C-1), 51.64 (C-7), 38.35 (C-10), 28.35 (C-8), 27.61 (C-11), 25.76
Chapter 6 : Experimental
164
(C-9), 22.42 (C-12)
HR-MS: m/z calculated for C13H20NaO4 [M+Na]+: 263.1254; found: 263.1256.
4-(Hydroxymethyl)-2-(4-methylpentyl)furan-3-carboxylic acid, AHFCA6 45
Following the general procedure for AHFCA methyl ester hydrolysis, work up gave
4-(hydroxymethyl)-2-(4-methylpentyl)furan-3-carboxylic acid (0.17 g, 95%) as a
colourless solid.
1H NMR (400 MHz, CDCl3) δ: 0.88 (d, 6H, H-11, J = 6.5 Hz), 1.23 (quartet, 2H, H-9, J
= 7.5 Hz), 1.57 (nonet, 1H, H-10, J = 6.5 Hz), 1.69 (quintet, 2H, H-8, J = 8.0 Hz), 2.98
(t, 2H, H-7, J = 7.5 Hz), 4.60 (s, 2H, H-1), 7.28 (s, 1H, H-3), O-H and COO-H proton
signal missing
13C NMR (100 MHz, CDCl3) δ: 169.97 (C-4), 167.14 (C-6), 138.70 (C-3), 125.45 (C-2),
111.30 (C-5), 55.63 (C-1), 38.38 (C-9), 28.43 (C-7), 27.72 (C-10), 25.75 (C-8), 22.48
(C-11)
HR-MS: m/z calculated for C12H18NaO4 [M+Na]+: 249.1097; found: 249.1108.
Melting point: 80 - 82 °C.
6.1.7.4 AHFCA7
Methyl 3-oxononanoate 128
Following the general procedure for β-ketoester synthesis, work up gave methyl
3-oxononanoate (0.35 g, 80%) as a colourless oil.
Chapter 6 : Experimental
165
1H NMR (300 MHz, CDCl3) δ: 0.84 (t, 3H, H-1, J = 6.5 Hz), 1.23 - 1.30 (m, 6H, H-2, 3
and 4), 1.55 (quintet, 2H, H-5, J = 7.0 Hz), 2.50 (t, 2H, H-6, J = 7.5 Hz), 3.42 (s, 2H, H-
8), 3.70 (s, 3H, H-10)
13C NMR (75 MHz, CDCl3) δ: 202.79 (C-7), 167.62 (C-9), 52.20 (C-10), 48.90 (C-8),
42.97 (C-6), 31.43 (C-3), 28.56 (C-4), 23.31 (C-5), 22.36 (C-2), 13.90 (C-1)
HR-MS: m/z calculated for C10H18NaO3 [M+Na]+: 209.1148, found: 209.1151.
Methyl 2-hexyl-4-(hydroxymethyl)furan-3-carboxylate 130
Following the general procedure for AHFCA cyclisation, work up gave methyl 2-hexyl-
4-(hydroxymethyl)furan-3-carboxylate (0.25 g, 74%) as a yellow oil.
1H NMR (300 MHz, CDCl3) δ: 0.87 (t, 3H, H-13, J = 6.5 Hz), 1.24 - 1.33 (m, 6H, H-10,
11 and 12), 1.63 (quintet, 2H, H-9, J = 7.5 Hz), 2.92 (t, 2H, H-8, J = 7.5 Hz), 3.51 (br s,
1H, O-H), 3.87 (s, 3H, H-7), 4.54 (s, 2H, H-1), 7.23 (s, 1H, H-3)
13C NMR (75 MHz, CDCl3) δ: 165.68 (C-4), 164.90 (C-6), 138.25 (C-3), 125.73 (C-2),
111.94 (C-5), 55.69 (C-1), 51.64 (C-7), 31.37 (C-8), 28.79 (C-11), 28.15 (C-9), 27.85
(C-10), 22.44 (C-12), 13.97 (C-13)
HR-MS: m/z calculated for C13H20NaO4 [M+Na]+: 263.1254; found: 263.1260.
Chapter 6 : Experimental
166
2-Hexyl-4-(hydroxymethyl)furan-3-carboxylic acid, AHFCA7 46
Following the general procedure for AHFCA methyl ester hydrolysis, work up gave
2-hexyl-4-(hydroxymethyl)furan-3-carboxylic acid (0.22 g, 95%) as a yellowish solid.
1H NMR (400 MHz, CDCl3) δ: 0.88 (t, 3H, H-12, J = 6.5 Hz), 1.25 - 1.36 (m, 6H, H-9,
10 and 11), 1.68 (quintet, 2H, H-8, J = 7.0 Hz), 3.00 (t, 2H, H-7, J = 7.5 Hz), 4.60 (s, 2H,
H-1), 7.28 (s, 1H, H-3), O-H and COO-H proton signal missing
13C NMR (100 MHz, CDCl3) δ: 170.17 (C-4), 167.17 (C-6), 138.68 (C-3), 125.43 (C-2),
111.33 (C-5), 55.61 (C-1), 31.44 (C-7), 28.87 (C-10), 28.23 (C-8), 27.82 (C-9), 22.49
(C-11), 14.04 (C-12)
HR-MS: m/z calculated for C12H18NaO4 [M+Na]+: 249.1097; found: 249.1108.
Melting point: 80-81 °C.
6.1.8 Synthesis of pyochelin and isopyochelin
(R)-2-(2-Hydroxyphenyl)-4,5-dihydrothiazole-4-carboxylic acid 131
A solution of 2-hydroxy-2-benzonitrile (1.90 g, 16.0 mmol) and L-cysteine (3.86 g,
31.9 mmol) in a 1:1 v/v mixture of methanol and phosphate buffer pH 6.4 (75 mL) was
stirred under argon at 50 °C for 4 days. The insoluble materials were removed by
filtration and the filtrate concentrated under vacuum. To the aqueous phase was added
20 mL water and 10 mL 1 M HCl, and the mixture extracted with DCM. The combined
Chapter 6 : Experimental
167
organic layers were washed with water and brine, dried over MgSO4 and concentrated
under vacuum. The residue was recrystallised using a mixture of ethanol and hexane to
yield (R)-2-(2-hydroxyphenyl)-4,5-dihydrothiazole-4-carboxylic acid (2.85 g, 80%) as a
yellow solid.
1H NMR (300 MHz, CDCl3): δ 3.61 - 3.73 (m, 2H, H-9), 5.42 (dd, 1H, H-8, J = 9.5 and
8.0 Hz), 6.86 - 6.92 (m, 1H, H-4), 7.03 (dd, 1H, H-2, J = 8.5 and 0.5 Hz), 7.35 - 7.39 (m,
1H, H-3), 7.43 (dd, 1H, H-1, J = 8.0 and 1.5 Hz), 10.79 (br s, 2H, O-H and COO-H)
13C NMR (75 MHz, DMSO-d6): δ 172.71 (C-10), 171.29 (C-7), 158.38 (C-5), 133.72
(C-3), 130.53 (C-1), 119.34 (C-2), 116.87 (C-4), 115.60 (C-6), 76.32 (C-8), 33.42 (C-9)
HR-MS: m/z calculated for C10H10NO3S [M+H]+: 224.0381, found: 224.0378.
Melting point: 114-116 °C.
2-(2-Hydroxyphenyl)-N-methoxy-N-methyl-4,5-dihydrothiazole-4-carboxamide
134
To a solution of (R)-2-(2-hydroxyphenyl)-4,5-dihydrothiazole-4-carboxylic acid (2.85 g,
12.8 mmol) in 30 mL THF was added CDMT (2.70 g, 15.4 mmol) and
N-methylmorpholine (4.2 mL, 38.2 mmol) at 0 °C and the mixture stirred for 1 hour,
followed by addition of N,O-dimethylhydroxylamine hydrochloride (1.25 g, 12.8 mmol).
The reaction mixture was then stirred at RT overnight and quenched with 50 mL water.
After removal of THF under vacuum, the remaining aqueous solution was acidified to
pH 4 and extracted with DCM. The combined organic layers were washed with water
and brine, dried over MgSO4 and concentrated under vacuum. The residue was purified
by column chromatography (pet ether/EtOAc = 2:1 v/v) to yield 2-(2-hydroxyphenyl)-
N-methoxy-N-methyl-4,5-dihydrothiazole-4-carboxamide (1.43 g, 42%) as a yellow oil.
1H NMR (300 MHz, CDCl3): δ 3.27 (s, 3H, H-11), 3.46 (dd, 1H, H-9a, J = 11.0 and
9.5 Hz), 3.75 (t, 1H, H-9b, J = 10.0 Hz), 3.81 (s, 3H, H-12), 5.68 (t, 1H, H-8, J= 9.0 Hz),
Chapter 6 : Experimental
168
6.86 (td, 1H, H-4, J = 7.5 and 1.0 Hz), 6.97 (dd, 1H, H-2, J = 8.5 and 1.0 Hz), 7.31 - 7.37
(m, 1H, H-3), 7.41 (dd, 1H, H-1, J =8.0 and 1.5 Hz), 12.32 (br s, 1H, O-H)
13C NMR (75 MHz, CDCl3): δ 173.98 (C-10), 169.62 (C-7), 158.87 (C-5), 133.25 (C-3),
130.71 (C-1), 118.86 (C-2), 116.96 (C-4), 116.03 (C-6), 74.51 (C-8), 61.72 (C-12), 32.76
(C-9), 32.45 (C-11)
HR-MS: m/z calculated for C12H14N2NaO3S [M+Na]+: 289.0617, found: 289.0618.
2-(2-Hydroxyphenyl)-4,5-dihydrothiazole-4-carbaldehyde 132
To a solution of 2-(2-hydroxyphenyl)-N-methoxy-N-methyl-4,5-dihydrothiazole-4-
carboxamide (0.45 g, 1.7 mmol) in 20 mL dry Et2O at -20 °C was added LiAlH4 (0.19 g,
5.0 mmol) and the reaction mixture stirred at the same temperature for 20 mins or until
the starting material had disappeared. Then 2 mL methanol, 20 mL saturated NH4Cl
solution and 10 mL 5% (v/v) H2SO4 solution were successively added. The organic phase
was conserved and the aqueous phase extracted with DCM. The combined organic layers
were washed with brine, dried over MgSO4 and concentrated under vacuum to yield the
crude 2-(2-hydroxyphenyl)-4,5-dihydrothiazole-4-carbaldehyde (0.21 g, 60%) as a
yellow oil. The aldehyde was unstable and therefore immediately used in the next step
without further purification.
HR-MS: m/z calculated for C10H10NO2S [M+H]+: 208.0427, found: 208.0432.
Pyochelin 74
Following the general condensation procedure with N-methyl-L-cysteine hydrochloride,
work up gave a mixture of pyochelin stereoisomers (29 mg, 74%) as a yellow solid.
Chapter 6 : Experimental
169
HR-MS: m/z calculated for C14H17N2O3S2 [M+H]+: 325.0675, found: 325.0681.
4"R-Isopyochelin 135
Following the general condensation procedure with 2-methyl-L-cysteine hydrochloride,
work up gave a mixture of 4"R-isopyochelin stereoisomers (25 mg, 64%) as a yellow
solid.
HR-MS: m/z calculated for C14H17N2O3S2 [M+H]+: 325.0675, found: 325.0677.
4"S-Isopyochelin 136
Following the general condensation procedure with 2-methyl-D-cysteine hydrochloride,
work up gave a mixture of 4"S-isopyochelin stereoisomers (31 mg, 79%) as a yellow
solid.
HR-MS: m/z calculated for C14H17N2O3S2 [M+H]+: 325.0675, found: 325.0674.
6.1.9 Synthesis of stereospecifically deuterium-labelled L-cystine
tert-Butyl (R)-(3-(benzyloxy)-1-(3,5-dimethyl-1H-pyrazol-1-yl)-1-oxopropan-2-
yl)carbamate 141
Chapter 6 : Experimental
170
To a solution of O-benzyl-N-(tert-butoxycarbonyl)-D-serine (7.60 g, 25.8 mmol),
3,5-dimethylpyrazole (2.72 g, 28.3 mmol) and DMAP (0.50 g, 4.1 mmol) in 40 mL
DCM was slowly added a solution of DCC (5.84 g, 28.3 mmol) in 10 mL DCM at 0 °C
and the mixture stirred at the same temperature for 0.5 hour. The reaction was then
allowed to warm up to RT and stirred overnight. After addition of 20 mL n-hexane, the
precipitated DCU was filtered to be removed. The filtrate was concentrated under
vacuum and the residue was purified by column chromatography (pet ether/EtOAc = 9:1
v/v) to yield tert-butyl (R)-(3-(benzyloxy)-1-(3,5-dimethyl-1H-pyrazol-1-yl)-1-
oxopropan-2-yl)carbamate (6.81 g, 71%) as a white solid.
1H NMR (300 MHz, CDCl3) δ: 1.45 (s, 9H, H-10), 2.16 (s, 3H, H-1), 2.53 (s, 3H, H-5),
3.83 (d, 1H, H-11a, J = 10.0 Hz), 4.05 (d, 1H, H-11b, J = 10.5 Hz), 4.44 (d, 1H, H-12a,
J = 12.5 Hz), 4.52 (d, 1H, H-12b, J = 12.5 Hz), 5.64 (br s, 2H, H-7 and N-H), 5.93 (s,
1H, H-3), 7.17 - 7.29 (m, 5H, H-14, 15 and 16)
13C NMR (75 MHz, CDCl3) δ: 170.38 (C-6), 155.41 (C-8), 152.35 (C-2), 144.19 (C-4),
137.61 (C-13), 128.18 (C-15), 127.50 (C-16), 127.34 (C-14), 111.12 (C-3), 79.71 (C-9),
72.79 (C-12), 70.31 (C-11), 54.14 (C-7), 28.23 (C-10), 14.12 (C-5), 13.65 (C-1)
HR-MS: m/z calculated for C20H28N3O4 [M+H]+: 374.2074; found: 374.2077.
Melting point: 66-67 °C.
tert-Butyl (R)-(1-(benzyloxy)-3-oxopropan-2-yl-3-2H)carbamate 142
To a solution of LiAlD4 (0.53 g, 12.3 mmol) in 12 mL THF was added a solution of tert-
butyl (R)-(3-(benzyloxy)-1-(3,5-dimethyl-1H-pyrazol-1-yl)-1-oxopropan-2-yl)
carbamate (2.30 g, 6.2 mmol) in 60 mL at -40 °C under argon and the mixture stirred for
1 hour. The reaction was quenched by adding 10% aqueous citric acid slowly and
allowed to warm up to RT. After removal of the insoluble materials by filtration through
celite, the filtrate was diluted with Et2O, washed with aqueous KHSO4 solution, dried
Chapter 6 : Experimental
171
over MgSO4 and concentrated to yield tert-butyl (R)-(1-(benzyloxy)-3-oxopropan-2-yl-
3-2H)carbamate (1.73 g, 100%) as a colourless oil. The crude product was used in the
next step without further purification.
1H NMR (400 MHz, CDCl3) δ: 1.44 (s, 9H, H-5), 3.69 (dd, 1H, H-6a, J = 9.5 and 3.5 Hz),
3.95 (dd, 1H, H-6b, J = 9.5 and 3.0 Hz), 4.29 (m, 1H, H-2), 4.45 (d, 1H, H-7a, J =
12.0 Hz), 4.50 (d, 1H, H-7b, J = 12.0 Hz), 5.53 (br d, 1H, N-H, J = 6.5 Hz), 7.24 - 7.34
(m, 5H, H-9, 10 and 11)
13C NMR (100 MHz, CDCl3) δ: 198.63 (t, C-1, J = 27.0 Hz), 155.44 (C-3), 137.13 (C-8),
128.27 (C-10), 127.73 (C-11), 127.47 (C-9), 79.90 (C-4), 73.25 (C-7), 67.60 (C-6), 59.74
(C-2), 28.08 (C-5)
HR-MS: m/z calculated for C15H20DNNaO4 [M+Na]+: 303.1426; found: 303.1430.
tert-Butyl ((2S,3R)-1-(benzyloxy)-3-hydroxypropan-2-yl-3-2H)carbamate 143
To a solution of tert-butyl (R)-(1-(benzyloxy)-3-oxopropan-2-yl-3-2H)carbamate (2.00 g,
7.1 mmol) in 70 mL THF was added a solution of 0.5 M S-Alpine-Borane in THF
(14.3 mL, 7.2 mmol) at RT under argon. After being stirred overnight, the reaction was
refluxed for 1 hour and cooled to RT. To the mixture was added acetaldehyde (0.8 mL)
and, after being stirred for 15 mins, the solvent was removed under vacuum. The residue
was dissolved in dry Et2O and to the solution was added diethanolamine (2.8 mL) at 0 °C
under argon. After being stirred for 30 mins, the precipitated solids were removed by
filtration and the filtrate was washed with water, dried over MgSO4 and concentrated.
The crude product was purified by column chromatography (pet ether/EtOAc = 2:1 v/v)
to yield tert-butyl ((2S,3R)-1-(benzyloxy)-3-hydroxypropan-2-yl-3-2H)carbamate
(1.92 g, 95%) as a colourless oil.
1H NMR (400 MHz, CDCl3) δ: 1.43 (s, 9H, H-5), 3.20 (br s, 1H, O-H), 3.56 (dd, 1H,
H-6a, J = 9.5 and 4.5 Hz), 3.61 (dd, 1H, H-6b, J = 9.0 and 4.0 Hz), 3.73 (s, 1H, H-1),
Chapter 6 : Experimental
172
3.78 (br s, 1H, H-2), 4.50 (s, 2H, H-7), 5.25 (s, 1H, N-H), 7.25 - 7.35 (m, 5H, H-9, 10
and 11)
13C NMR (100 MHz, CDCl3) δ: 155.91 (C-3), 137.64 (C-8), 128.28 (C-10), 127.62
(C-11), 127.47 (C-9), 79.39 (C-4), 73.17 (C-7), 69.93 (C-6), 62.64 (t, C-1, J = 22.0 Hz),
51.40 (C-2), 28.21 (C-5)
HR-MS: m/z calculated for C15H22DNNaO4 [M+Na]+: 305.1575; found: 305.1582.
(1R,2R)-3-(Benzyloxy)-2-((tert-butoxycarbonyl)amino)propyl-1-2H acetate 144
To a solution of tert-butyl ((2S,3R)-1-(benzyloxy)-3-hydroxypropan-2-yl-3-
2H)carbamate (2.55 g, 9.0 mmol) in 20 mL DCM was added triethylamine (3.8 mL,
27.1 mmol), acetic anhydride (1.40 g, 13.6 mmol) and DMAP (0.33 g, 2.7 mmol) and
the mixture stirred at RT for 5 hours. After removal of the solvent, 30 mL water was
added and the mixture extracted with Et2O. The combined organic phase was washed
with aqueous KHSO4 solution, dried over MgSO4 and concentrated under vacuum to
yield (1R,2R)-3-(benzyloxy)-2-((tert-butoxycarbonyl)amino)propyl-1-2H acetate (2.90 g,
99%) as a colourless oil.
1H NMR (400 MHz, CDCl3) δ: 1.44 (s, 9H, H-7), 2.00 (s, 3H, H-1), 3.49 (dd, 1H, H-8a,
J = 9.5 and 5.0 Hz), 3.55 (dd, 1H, H-8b, J = 9.5 and 3.0 Hz), 4.03 (br s, 1H, H-4), 4.17
(d, 1H, H-3, J = 6.0 Hz), 4.48 (d, 1H, H-9a, J = 12.0 Hz), 4.52 (d, 1H, H-9b, J = 12.0 Hz),
5.06 (br d, 1H, N-H, J = 8.0 Hz), 7.27 - 7.35 (m, 5H, H-11, 12 and 13)
13C NMR (100 MHz, CDCl3) δ: 170.64 (C-2), 155.25 (C-5), 137.55 (C-10), 128.16
(C-12), 127.53 (C-13), 127.43 (C-11), 79.36 (C-6), 72.95 (C-9), 68.50 (C-8), 63.02 (t,
C-3, J = 22.5 Hz), 48.77 (C-4), 28.07 (C-7), 20.50 (C-1)
HR-MS: m/z calculated for C17H24DNNaO5 [M+Na]+: 347.1688; found: 347.1700.
Chapter 6 : Experimental
173
(1R,2R)-2-((tert-Butoxycarbonyl)amino)-3-hydroxypropyl-1-2H acetate 145
A mixture of (1R,2R)-3-(benzyloxy)-2-((tert-butoxycarbonyl)amino)propyl-1-2H acetate
(6.30 g, 19.4 mmol) and 10% palladium on carbon (1.80 g) in 120 mL methanol was
hydrogenated at RT for 2 hours. The catalyst was removed by filtration through celite
and the filtrate was concentrated to give (1R,2R)-2-((tert-butoxycarbonyl)amino)-3-
hydroxypropyl-1-2H acetate (4.44 g, 98%) as a colourless oil.
1H NMR (400 MHz, CDCl3) δ: 1.44 (s, 9H, H-7), 2.08 (s, 3H, H-1), 3.63 (dd, 1H, H-8a,
J =11.5 and 5.0 Hz), 3.68 (dd, 1H, H-8b, J = 11.5 and 4.5 Hz), 3.88 (br s, 1H, H-4), 4.15
(d, 1H, H-3, J = 5.5 Hz), 4.57 (br s, 1H, O-H), 5.49 (br d, 1H, N-H, J = 7.0 Hz)
13C NMR (100 MHz, CDCl3) δ: 171.07 (C-2), 155.72 (C-5), 79.33 (C-6), 62.68 (t, C-3,
J = 22.5 Hz), 61.00 (C-8), 50.56 (C-4), 27.91 (C-7), 20.36 (C-1)
HR-MS: m/z calculated for C10H18DNNaO5 [M+Na]+: 257.1218; found: 257.1215.
(3R)-O-Acetyl-N-(tert-butoxycarbonyl)-L-serine-3-2H 146
To a suspension of NaIO4 (48.3 g, 225.7 mmol) and RuCl3·nH2O (1.40 g) in 115 mL
water in an ice bath was slowly added a solution of (1R,2R)-2-((tert-
butoxycarbonyl)amino)-3-hydroxypropyl-1-2H acetate (4.44 g, 19.0 mmol) in 115 mL
acetone. Then the reaction mixture was allowed to stir at RT for 1.5 hours. To the
reaction was added isopropanol (60 mL) and the mixture stirred in an ice bath for 1 hour.
The precipitated RuO2 was removed by filtration through celite and the filtrate
concentrated under vacuum to remove the acetone and extracted with chloroform. The
Chapter 6 : Experimental
174
combined chloroform layers were washed with brine, dried over MgSO4 and
concentrated to yield (3R)-O-acetyl-N-(tert-butoxycarbonyl)-L-serine-3-2H (4.70 g,
100%) as a colourless oil.
1H NMR (400 MHz, CDCl3) δ: 1.46 (s, 9H, H-7), 2.08 (s, 3H, H-1), 4.43 (d, 1H, H-3, J
=3.5 Hz), 4.62 (dd, 1H, H-4, J = 8.0 and 4.0 Hz), 5.56 (br d, 1H, N-H, J = 8.0 Hz), 11.00
(s, 1H, O-H)
13C NMR (100 MHz, CDCl3) δ: 172.96 (C-8), 170.80 (C-2), 155.35 (C-5), 80.38 (C-6),
63.83 (t, C-3, J = 22.0 Hz), 52.43 (C-4), 28.01 (C-7), 20.39 (C-1)
HR-MS: m/z calculated for C10H16DNNaO6 [M+Na]+: 271.1011; found: 271.1009.
tert-Butyl (R)-O-acetyl-N-(tert-butoxycarbonyl)-L-serinate-3-2H 147
To a refluxing solution of (3R)-O-acetyl-N-(tert-butoxycarbonyl)-L-serine-3-d (4.90 g,
19.8 mmol) in 75 mL toluene was added a mixture of tert-butyl alcohol (5.97 g,
80.6 mmol) and N,N-dimethylformamide dineopentyl acetal (12.44 g, 53.8 mmol) and
the reaction mixture refluxed for 6.5 hours. Then the cooled reaction was diluted with
EtOAc, washed with saturated NaHCO3 solution and brine, and dried over MgSO4. The
organic layer was concentrated under vacuum and the residue purified by column
chromatography (pet ether/EtOAc = 9:1 v/v) to yield tert-butyl (R)-O-acetyl-N-(tert-
butoxycarbonyl)-L-serinate-3-2H (4.50 g, 75%) as a colourless oil.
1H NMR (400 MHz, CDCl3) δ: 1.35 (s, 9H, H-7), 1.36 (s, 9H, H-10), 1.94 (s, 3H, H-1),
4.32 - 4.34 (m, 2H, H-3 and 4), 5.28 (br d, 1H, N-H, J = 7.0 Hz)
13C NMR (100 MHz, CDCl3) δ: 170.17 (C-8), 168.49 (C-2), 155.00 (C-5), 82.34 (C-9),
79.70 (C-6), 64.13 (t, C-3, J = 23.5 Hz), 53.17 (C-4), 28.06 (C-10), 27.66 (C-7), 20.36
(C-1)
HR-MS: m/z calculated for C14H25DNO6 [M+H]+: 305.1817; found: 305.1824.
Chapter 6 : Experimental
175
tert-Butyl (R)-(tert-butoxycarbonyl)-L-serinate-3-2H 148
To a solution of tert-butyl (R)-O-acetyl-N-(tert-butoxycarbonyl)-L-serinate-3-2H (4.20 g,
13.8 mmol) in 65 mL THF was added a solution of LiOH (0.50 g, 20.7 mmol) in 50 mL
water and the reaction stirred at RT for 3 hours. The mixture was extracted with EtOAc
and the combined organic layers washed successively with water and brine and dried
over MgSO4. After removal of the solvent under vacuum, the residue was purified by
column chromatography (pet ether/EtOAc = 4:1 v/v) to yield tert-butyl (R)-(tert-
butoxycarbonyl)-L-serinate-3-2H (2.24 g, 62%) as a colourless oil.
1H NMR (400 MHz, CDCl3) δ: 1.36 (s, 9H, H-5), 1.39 (s, 9H, H-8), 3.23 (br s, 1H, O-H),
3.79 (br s, 1H, H-1), 4.14 (br s, 1H, H-2), 5.52 (br d, 1H, N-H, J = 7.0 Hz)
13C NMR (100 MHz, CDCl3) δ: 169.80 (C-6), 155.73 (C-3), 82.08 (C-7), 79.71 (C-4),
63.02 (t, C-1, J = 23.0 Hz), 56.09 (C-2), 28.11 (C-8), 27.78 (C-5)
HR-MS: m/z calculated for C12H22DNNaO5 [M+Na]+: 285.1531; found: 285.1538.
tert-Butyl (R)-N-(tert-butoxycarbonyl)-O-tosyl-L-serinate-3-2H 149
To a solution of tert-butyl (R)-(tert-butoxycarbonyl)-L-serinate-3-2H (2.24 g, 8.5 mmol)
in 10 mL pyridine was added 4-toluenesulfonyl chloride (3.00 g, 15.7 mmol) at 0 °C and
the reaction stirred at 4 °C overnight. The mixture was quenched with 10 mL water and
extracted with EtOAc. The combined organic layers were washed successively with 1 M
HCl, water and brine, and dried over MgSO4. After removal of the solvent under vacuum,
the residue was purified by column chromatography (pet ether/EtOAc = 6:1 v/v) to yield
Chapter 6 : Experimental
176
tert-butyl (R)-N-(tert-butoxycarbonyl)-O-tosyl-L-serinate-3-2H (3.10 g, 89%) as a white
solid.
1H NMR (400 MHz, CDCl3) δ: 1.41 (s, 9H, H-10), 1.43 (s, 9H, H-13), 2.44 (s, 3H, H-1),
4.35 - 4.38 (m, 2H, H-6 and 7), 5.27 (br d, 1H, N-H, J = 7.5 Hz), 7.35 (d, 2H, H-3, J =
8.0 Hz), 7.77 (d, 2H, H-4, J = 8.5 Hz)
13C NMR (100 MHz, CDCl3) δ: 167.29 (C-11), 154.91 (C-8), 144.97 (C-2), 132.42 (C-5),
129.83 (C-3), 127.88 (C-4), 83.22 (C-12), 80.07 (C-9), 69.60 (t, C-6, J = 23.5 Hz), 53.21
(C-7), 28.14 (C-13), 27.73 (C-10), 21.54 (C-1)
HR-MS: m/z calculated for C19H29DNO7S [M+H]+: 417.1800; found: 417.1807.
Melting point: 91 - 92 °C.
tert-Butyl (S)-S-acetyl-N-(tert-butoxycarbonyl)-L-cysteinate-3-2H 150
To a solution of tert-butyl (R)-N-(tert-butoxycarbonyl)-O-tosyl-L-serinate-3-2H (0.80 g,
1.9 mmol) in 5 mL DMF was added potassium thioacetate (0.43 g, 3.8 mmol) under
argon at 0 °C and the reaction stirred at RT overnight. The solvent was removed under
vacuum and the residue extracted with EtOAc. The combined organic phase was washed
with water and brine, dried over MgSO4 and concentrated under vacuum. The crude
product was purified by column chromatography (pet ether/EtOAc = 6:1 v/v) to yield
tert-butyl (S)-S-acetyl-N-(tert-butoxycarbonyl)-L-cysteinate-3-2H (0.61 g, 99%) as a
white solid.
1H NMR (400 MHz, CDCl3) δ: 1.29 (s, 9H, H-7), 1.32 (s, 9H, H-10), 2.19 (s, 3H, H-1),
3.25 (br d, 1H, H-3, J = 4.0 Hz), 4.24 (br dd, 1H, H-4, J = 8.0 and 4.5 Hz), 5.21 (br d,
1H, N-H, J = 7.5 Hz)
Chapter 6 : Experimental
177
13C NMR (100 MHz, CDCl3) δ: 193.95 (C-2), 169.06 (C-8), 154.78 (C-5), 82.12 (C-9),
79.36 (C-6), 53.26 (C-4), 30.75 (t, C-3, J = 22.0 Hz), 30.10 (C-1), 27.94 (C-10), 27.52
(C-7)
HR-MS: m/z calculated for C14H24DNNaO5S [M+Na]+: 343.1408; found: 343.1400.
Melting point: 51-53 °C.
(2R,2'R,3S,3'S)-[3,3'-2H2]Cystine 151
A solution of tert-butyl (S)-S-acetyl-N-(tert-butoxycarbonyl)-L-cysteinate-3-2H (0.61 g,
1.9 mmol) in 20 mL of 6 M HCl was refluxed overnight. The cooled reaction mixture
was washed with chloroform and concentrated. The residue was treated with prewashed
and activated ion-exchange resin (Dowex® 50WX8 hydrogen form, 200-400 mesh) and
the adsorbed amino acid eluted with 1 M ammonia solution in methanol. The fractions,
which gave a positive ninhydrin reaction, were combined, aerated for 48 hours, and
concentrated to give (2R,2'R,3S,3'S)-[3,3'-2H2]cystine (0.18 g, 78%) as a white solid.
1H NMR (300 MHz, 2.5% NaOD-D2O) δ: 2.97 (d, 1H, H-3, J = 3.5 Hz), 3.45 (d, 1H,
H-2, J = 3.5 Hz)
13C NMR (125 MHz, 2.5% NaOD-D2O) δ: 180.81 (C-1), 54.80 (C-2), 43.30 (t, C-3, J =
22.0 Hz)
HR-MS: m/z calculated for C6H11D2N2O4S2 [M+H]+: 243.0437; found: 243.0440.
Melting point: > 200 °C dec. (open capillary) The sample showed slight discolorization
at 190 °C and turned black from 200 °C.
Chapter 6 : Experimental
178
6.2 Analytical chemistry
6.2.1 Extraction protocol
For extracts from liquid bacterial cultures, ethanol was added (in equal volume to the
volume of culture media used) and the mixture shaken for 1 hour (180 rpm, 30 °C) before
the insoluble materials were pelleted at 13200 rpm for 10 mins. The supernatant was then
filtered (0.4 μm) for LC-MS analyses.
6.2.2 Liquid chromatography-mass spectrometry (LC-MS)
Unless otherwise stated, LC-MS analyses were performed on a Dionex Ultimate 3000
HPLC instrument equipped with a ZORBAX Eclipse Plus C18 column (2.1 × 100 mm,
1.8 μm) coupled to a Bruker MaXis Impact mass spectrometer [ESI in positive ion mode;
full scan 50-2500 m/z; end plate offset, -500 V; capillary, -4500 V; nebulizer gas (N2),
1.4 bar; dry gas (N2), 8 L/min; dry temperature, 200 °C]. The solvents used for elution
of the column were as follows: solvent A, 0.1% formic acid in water; solvent B, 0.1%
formic acid in acetonitrile. The elution profile was as follows: 0-5 min, 5% B; 5-17 min,
linear gradient from 5% B to 100% B; 17-22 min, 100% B; 22-25 min, linear gradient
from 100% B to 5% B; 25-34 min, 5% B. The flow rate was 0.2 mL/min. The mass
spectrometer was calibrated with 10 mM sodium formate at the beginning of each run.
For comparison of MS/MS fragmentation pattern of synthetic butenolide and the
dephosphorylated product of the MmfL-catalysed reaction, LC-MS/MS analyses were
performed on an Agilent 1260 Infinity HPLC equipped with an Agilent Eclipse XDB-
C18 column (4.6 × 150 mm, 5 μm) coupled to a Bruker AmaZon X ion trap mass
spectrometer [ESI in positive ion mode; full scan 100-3000 m/z; end plate offset, -500 V;
capillary, -4500 V; nebulizer gas (N2), 40 p.s.i.; dry gas (N2), 10 L/min; dry temperature,
200 °C]. The solvents used for elution of the column were as follows: solvent A, 0.1%
formic acid in water; solvent B, 0.1% formic acid in acetonitrile. The elution profile was
as follows: 0-5 min, 5% B; 5-30 min, linear gradient from 5% B to 100% B; 30-35 min,
100% B; 35-40 min, linear gradient from 100% B to 5% B. The flow rate was 1.0 mL/min.
For protein samples (25-45 kDa), they are diluted in distilled water to a final
concentration of 3-4 μM. LC-MS analyses were performed on a Dionex Ultimate 3000
HPLC instrument equipped with an ACE C4-300 column (2.1 × 100 mm, 5 μm) coupled
Chapter 6 : Experimental
179
to a Bruker MaXis II mass spectrometer [ESI in positive ion mode; full scan 200-
3000 m/z; end plate offset, -500 V; capillary, -4500 V; nebulizer gas (N2), 1.8 bar; dry
gas (N2), 9 L/min; dry temperature, 200 °C]. The solvents used for elution of the column
were as follows: solvent A, 0.1% formic acid in water; solvent B, 0.1% formic acid in
acetonitrile. The elution profile was as follows: 0-2.5 min, 5% B; 2.5-13.4 min, linear
gradient from 5% B to 100% B; 13.4-18.4 min, 100% B; 18.4-21.4 min, linear gradient
from 100% B to 5% B; 21.4-30 min, 5% B. The flow rate was 0.2 mL/min.
6.2.3 Chiral LC-MS analysis
An Agilent 1260 Infinity HPLC coupled to a Bruker AmaZon X ion trap mass
spectrometer [ESI in positive ion mode; full scan 100-3000 m/z; end plate offset, -500 V;
capillary, -4500 V; nebulizer gas (N2), 10 p.s.i.; dry gas (N2), 4 L/min; dry temperature,
180 °C] was used for isopyochelin samples on a homochiral stationary phase. Synthetic
isopyochelins and the ethanol extract of culture broth from S. coelicolor M1152/SV-
2_E03::SspI were analysed on a ChiralPAK IE analytical column (4.6 × 150 mm, 5 μm)
in reverse phase mode. The solvents used for elution of the column were as follows:
solvent A, 0.1% formic acid in water; solvent B, 0.1% formic acid in acetonitrile. The
elution profile was as follows: 0-5 min, 20% B; 5-30 min, linear gradient from 20% B to
100% B; 30-35 min, 100% B; 35-40 min, linear gradient from 100% B to 20% B. The
flow rate was 0.8 mL/min.
6.2.4 Circular dichroism (CD) spectroscopy
The instrument used was a Jasco J-815 CD Spectrometer for analyses of purified MmfR
and its mutant proteins as well as the ordered 2-methyl-L-cysteine and 2-methyl-D-
cysteine. Concentrated proteins were diluted in 5 mM phosphate buffer pH 7.0 to a final
concentration of 0.1 or 0.01 mg/mL and measured in the range from 180 to 260 nm.
Chemical samples were dissolved in distilled water to a final concentration of 1 mg/mL
and measured in the range from 180 to 300 nm. The samples were loaded on a 1 mm path
length cuvette and instrument settings were as follows: sensitivity standard, data pitch
0.2 nm, bandwidth 1.0 nm, data integration time (D.I.T) 1 s, scanning speed 100 nm/min,
accumulation 4. Each spectrum was calibrated by subtracting a reference spectrum from
a sample containing only buffer or water.
Chapter 6 : Experimental
180
6.3 Biology
6.3.1 Bacterial strains
The wild type S. coelicolor M1152 strain was obtained from Dr. Vincent Poon. Strains
of S. coelicolor M1152/SV-2_E03::SspI and deletion mutants ∆sven0508, ∆sven0515
and ∆sven0516 were constructed by Dr. Yuki Inahashi in the Challis group.
6.3.2 Culture media
Mannitol soya flour medium (SFM) was prepared by dissolving Bacto agar (20 g),
mannitol (20 g) and soya flour (20 g) in tap water (1 L) and autoclaving before use.
Yeast-extract dextrose (YD) liquid medium was prepared by dissolving yeast extract
(10 g) and D-glucose (10 g) in distilled water (1 L) and autoclaving before use.
LB medium was prepared using LB Broth, Miller (Fisher) to a final concentration of
25 g/L. For solid LB, Bacto agar was added (15 g/L).
Supplemented minimal medium (SMM) was made up with the following ingredients
added to the concentrations as follows: Bacto casaminoacids (2 g) and TES buffer (5.73 g)
were dissolved in distilled water (1 L), the pH adjusted to 7.2 using 10 N NaOH and
autoclaved. At the time of use, the following ingredients were added: NaH2PO4 +
K2HPO4 (50 mM each, autoclaved, 10 mL per litre of culture), MgSO4 (1 M, autoclaved,
5 mL per litre of culture), D-glucose (50% w/v, autoclaved, 18 mL per litre of culture)
and trace elements solution (1 mL per litre of culture). Trace elements solution contained
0.1 g each of ZnSO4.7H2O, FeSO4.7H2O, MnCl2.4H2O, CaCl2.6H2O and NaCl in
distilled water (1 L). The solution was stored at 4 °C and refreshed every 4 weeks.
6.3.3 Preparation of bacterial spores
Mannitol SFM was used for the production of bacterial spores. Media was melted and
apramycin added to a final concentration of 50 μg/mL for growth of the S. coelicolor
M1152/SV-2_E03::SspI and ∆sven0508, ∆sven0515 and ∆sven0516 mutant strains.
Each plate of SFM was inoculated with 20 μL of bacterial spore stocks. After incubating
at 30 °C for 1 week, 5 mL sterile water was added to each plate and an inoculation loop
was used to scrape the spores on the surface. The spores suspended in the water were
Chapter 6 : Experimental
181
collected and filtered through a non-absorbent cotton wool using a syringe to remove
mycelia. The suspension was then centrifuged at 3000 rpm for 10 mins to pellet the
spores and the supernatant removed, leaving approximately 1 mL of suspension. Spore
suspension was transferred to a cryotube and an equal volume of sterile 50% glycerol
was added and the resulting spore stocks in 25% glycerol stored at -80 °C for future use.
6.3.4 Production of 2-hydroxyphenylthiazoline metabolites
Unless otherwise stated, strains of S. coelicolor M1152/SV-2_E03::SspI and deletion
mutants ∆sven0508, ∆sven0515 and ∆sven0516 were grown (30 °C, 180 rpm) in liquid
YD medium for production of the metabolites. Seed cultures were prepared in 50 mL of
YD medium, inoculated with 20 μL of spores. After incubation for 2 days, 2% of this
culture was used to inoculate 250-mL Erlenmeyer flasks, each containing 50 mL of YD
medium. The resulting culture was grown for 5 days. For analysis of these cultures,
samples were prepared for LC-MS analyses as described in Section 6.2.1 and 6.2.2.
6.3.5 Feeding the extract from ∆sven0515 mutant to ∆sven0516 mutant
50 mL of the 5-day-old culture of the ∆sven0515 mutant was first centrifuged at
5,000 rpm for 10 mins to remove the mycelium. Then the supernatant was acidified with
1 M HCl to pH 5.0 and extracted with EtOAc. The EtOAc layers were combined, dried
over MgSO4 and concentrated under vacuum to give the residue containing the
metabolites from the culture of the ∆sven0515 mutant. The extract was dissolved in 3 mL
of YD medium and fed into a 50 mL culture of the ∆sven0516 mutant which was
incubated (30 °C, 180 rpm) for 1 day after inoculation with 2% of the corresponding seed
culture. The resulting culture was grown for 4 days and prepared for LC-MS analyses as
described in Section 6.2.1 and 6.2.2.
6.3.6 Feeding experiment with deuterium-labelled precursors
50 mL of YD medium was inoculated with 20 μL spores of S. coelicolor
M1152/SV-2_E03::SspI and incubated (30 °C, 180 rpm) for 2 days. Then the cells were
pelleted at 3000 rpm for 5 mins, washed once with SMM medium and resuspended in
50 mL of SMM medium. 10 mL of SMM medium was inoculated with 0.2 mL (2% v/v)
of this bacterial suspension and incubated (30 °C, 180 rpm). After 24 hours, a suspension
Chapter 6 : Experimental
182
of precursor (2R,2'R,3S,3'S)-[3,3'-2H2]cystine or 2,3,3-2H3-L-cysteine (CK Isotopes) in
sterile water was added to give a final concentration of 1.25 mM. In some instances, a
solution of L-methionine (MP Biomedicals) was also added at the same time. This
feeding procedure was repeated at 24 hour intervals until a total incubation time of 5 days,
after which the metabolites were extracted and analysed as described in Section 6.2.1
and 6.2.2.
6.3.7 Chemical transformations
For subsequent transformations, 50 μL of chemically competent E. coli TOP10
(Invitrogen), E. coli BL21 star (DE3) (Invitrogen) or E. coli BL21 star (DE3)pLysS
(Invitrogen) cells suspension was incubated on ice for 30 mins after a gentle mixing with
1 μL plasmid. The cells were heat-shocked in a 42 °C water bath for 30 seconds and
immediately transferred on ice. 250 μL of room temperature LB (or SOC) medium was
added and cells incubated (37 °C, 180 rpm) for 1 hour before being spread in two
different portions (50 and 200 μL) on LB plates containing 100 μg/mL of ampicillin and
incubated at 37 °C for 16 hours. Single colonies were picked and incubated further in
10 mL LB containing ampicillin (37 °C, 180 rpm) for 16 hours. 0.6 mL of each culture
was taken and added to 0.6 mL of sterile 50% glycerol for storage at -80 °C.
For transformations using the E. coli TOP10 cells (IBA StarGate), most of the procedures
were same as above except the following modifications according to the manufacturer’s
protocol: cells were heat-shocked at 37 °C for 5 mins, and 900 mL of LB medium was
used for the 45-min pre-incubation before spreading the cells onto selection plates.
6.3.8 Plasmid DNA preparation
10 μL of glycerol stock of E. coli TOP10 cells carrying the plasmid of interest was used
to inoculate 10 ml LB medium containing 100 μg/mL of ampicillin. The liquid culture
was shaken for 16 hours (37 °C, 180 rpm). The culture was centrifuged at 2000 rpm for
10 mins and the supernatant discarded. GeneJET plasmid miniprep kit (Thermo
Scientific) was used for isolation of the plasmid DNA as per the manufacturer’s
guidelines. The pelleted cells were resuspended in 250 μL of the resuspension solution
by vortexing. The cells were lysed by adding 250 μL of the lysis solution and mixed by
inversion. 350 μL of the neutralization solution was then added before centrifugation to
remove the cell debris and chromosomal DNA. The supernatant was loaded onto a
Chapter 6 : Experimental
183
GeneJET spin column and centrifuged before the adsorbed plasmid DNA was washed
with 500 μL of the wash solution twice. The plasmid DNA was eluted by addition of
50 μL of sterile water on the purification column membrane and centrifuged after
incubation at RT for 2 mins. The concentration of purified plasmid DNA solution was
determined using a NanoDrop Lite spectrophotometer (Thermo Scientific). The plasmid
DNA was stored at -20 °C.
6.3.9 Restriction digests
Unless manufacturer’s guidelines recommended otherwise, 0.5 μL of restriction
endonuclease enzyme was incubated with 2 μL of the recommended buffer and 15 μL of
the prepared plasmid DNA at 37 °C for 2 hours, after which 4 μL of 6X DNA loading
dye was added to the reaction mixture and an agarose gel run (1%, w/v) of the resulting
fragments. Reactions using FastDigest enzymes were incubated for 30 mins.
6.3.10 Polymerase Chain Reaction (PCR)
An Eppendorf Mastercycler gradient and an Eppendorf Mastercycler personal PCR
machine were used for variable annealing temperature and constant annealing
temperature PCR reactions, respectively. Unless otherwise stated, the following
quantities of each component were used per reaction: sterile water (35.5 μL), DMSO
(5 μL), MgCl2 buffer (5 μL), dNTPs (2.5 mM each, 1 μL), Fermentas Taq DNA
polymerase (5 U/μL, 0.5 μL), DNA template (1 μL). 1 μL of each forward and reverse
primers (10 μM) were used for each reaction. The DNA template was about 100 ng of
cosmid used in each reaction, the volume of which was adjusted according to its
concentration. The volume of sterile water was adjusted accordingly to make a total
volume of 50 μL.
Cosmid C73-787 was used as the template for amplification of small DNA sequences
containing the operators for MmfR and provided by Dr. Christophe Corre. Cosmid
SV4 F01 was used as the template for cloning of sgnR gene and amplification for the
intergenic region containing the operator for SgnR and prepared by Dr. John D. Sidda.
Primers were obtained from Sigma-Genosys and diluted to 100 μM as per the
manufacturer’s guidelines before use. The typical program below was used for the single
annealing temperature PCR reactions: Reaction mixture heated to 95 °C (5 mins),
Chapter 6 : Experimental
184
35 cycles consisting of denaturation at 95 °C (1 min), annealing (1 min) and elongation
at 72 °C. This was followed by a further elongation step at 72 °C for 10 mins. The
reaction(s) were then held at 4 °C until PCR products were used for agarose gel
electrophoresis. The annealing temperature and length of time for elongation were
determined based on the primers used and length of the target DNA sequence. The
specific conditions and primers used in this study was summarised in Table 6.1.
Table 6.1. Primers and PCR conditions used in this study.
Target Primers Annealing
temp.
Elongation
time Purpose
sgnR
(600 bp)
For 5’-CACCGCGACACCCCGAAGCCAA-3’
Rev 5’-TCAGCGGCGCAGGCC-3’ 58 °C 50 s
Producing PCR product for
cloning of sgnR gene into
pET151/D-TOPO vector
sgnR-sgnL
(121 bp)
For 5’-GCCTGTCCGCCTCTCTCGTTC-3’
Rev 5’-CCTTTTCCCCCTCAAGGTGAATC-3’ 60 °C 20 s
Entire intergenic region between
sgnR and sgnL containing the
proposed binding site for SgnR
mmyR
promoter
(150 bp)
For 5’-ATCCTGCCGCGCGGTAGCC-3’
Rev 5’-CAACGCCCGAGTCCTCTCAAG-3’ 53 °C 20 s
Entire intergenic region near
mmyR promoter containing the
proposed binding site for MmfR
mmyR
promoter
(100 bp)
For 5’-GACATCGTCCACGTGCGCCG-3’
Rev 5’-CAACGCCCGAGTCCTCTCAAG-3’ 53 °C 20 s
Partial sequence near mmyR
promoter containing the
proposed binding site for MmfR
mmfR-mmfL
(194 bp)
For 5’-GGCTGCCTTCCTTCGTGTG-3’
Rev 5’-AGGGGCGCTACATCTCCCG-3’ 50 °C 30 s
Entire intergenic region between
mmfR and mmfL containing the
proposed binding site for MmfR
mmfR-mmfL
(100 bp)
For 5’-CACGGAAACCCATTGCATAATACC-3’
Rev 5’-GGACCACCGGCTGGCTTGC-3’ 55 °C 20 s
Partial sequence between mmfR
and mmfL containing the
proposed binding site for MmfR
mmyY-mmyB
(230 bp)
For 5’-GGTGAACTCCTTCGGCGAG-3’
Rev 5’-GGCGCCTCACAGTGTCAAAC-3’ 55 °C 20 s
Entire intergenic region between
mmyY and mmyB containing the
proposed binding site for MmfR
mmyY-mmyB
(98 bp)
For 5’-TCCAAACACCGAGGCCCCG-3’
Rev 5’-GGCGCCTCACAGTGTCAAAC-3’ 55 °C 20 s
Partial sequence between mmyY
and mmyB containing the
proposed binding site for MmfR
6.3.11 Agarose gel electrophoresis
Agarose electrophoresis gel (1%, w/v) was prepared by mixing 1 g of LE agarose
(SeaKem) and 100 mL of 1X TBE buffer. TBE buffer contained Tris base (10.8 g/L),
boric acid (5.5 g/L), and EDTA (1.17 g/L). The mixture was heated in a microwave until
complete dissolution of the agarose. After the gel mixture cooled down, 5 μL GelRed
was added and the gel poured into the gel plate with appropriate gel comb. The gel was
allowed to cool further until it set.
Chapter 6 : Experimental
185
To 5 μL DNA sample, 1 μL of 6X DNA loading dye (New England Biolabs) was added
and the samples loaded onto the prepared agarose gel for electrophoresis at 80 V using a
BioRad tank and Power PAC 300 with a UVP imaging system UV transilluminator.
6.3.12 Purification of DNA from agarose gels
After the electrophoresis, the bands of interest were cut out and placed in Eppendorf
tubes. GeneJET gel extraction kit (Thermo Scientific) was used to extract the DNA from
the excised gel slice as per the manufacturer’s guidelines. One volume of binding buffer
to one volume of gel slice was added and the mixture was incubated at 55 °C until melted
(10 mins + shaking every 2 mins). For DNA fragment whose size is no more than 500 bp,
one volume of isopropanol was also added to the sample. After mixing thoroughly by
vortexing, the DNA was then extracted using a GeneJET purification column and washed
with the wash buffer before the DNA was eluted by addition of 50 μL of sterile water on
the purification column membrane and centrifuged. The concentration of purified DNA
solution was determined using a NanoDrop Lite spectrophotometer (Thermo Scientific).
The DNA was stored at -20 °C.
6.3.13 Construction of plasmid pET151-sgnR
The sgnR gene was amplified by PCR with appropriate primers and conditions as
described in Section 6.3.10 (Table 6.1). PCR products were run on a 1% agarose gel, the
bands of interest excised, and purified using the GeneJET Gel Extraction Kit as described
in Section 6.3.12.
The cloning reaction was carried out according to the Invitrogen TOPO cloning manual.
Concentration of the PCR product was determined using a NanoDrop Lite
spectrophotometer (Thermo Scientific), and diluted accordingly to fall within the desired
range of 1-5 ng/μL for cloning into pET151 vector. To a 6 μL reaction, 1 μL of the
diluted PCR product, 1 μL of pET151/D-TOPO vector, 1 μL of the provided salt solution
and sterile water were mixed together gently and incubated for 5 mins at room
temperature and then placed on ice. E. coli TOP10 cells (IBA StarGate) were transformed
with 3 μL of the TOPO cloning reaction mixture as described in Section 6.3.7 and poured
onto selection LB plates containing 100 μg/mL of ampicillin.
Chapter 6 : Experimental
186
Colonies (typically 5 to 10) were picked from these plates, added to 10 mL LB medium
containing 100 μg/mL of ampicillin and incubated at 37 °C for 16 hours. Glycerol stocks
were prepared and plasmids purified from these overnight cultures as described in
Section 6.3.8. The purified plasmid was confirmed as containing the correct sgnR gene
insert by PCR with the primers same as above and by restriction digests. Plasmids that
gave both the expected sized PCR product and expected sized DNA fragments resulting
from the restriction digests were sent for sequencing by GATC Biotech, during which
the gene of interest was sequenced using the T7 primer pairs: forward 5’-
TAATACGACTCACTATAGGG-3’ and reverse 5’-CTAGTTATTGCTCAGCGG-3’.
6.3.14 Site-directed mutagenesis of MmfR
The construct pET151/D-TOPO vector with MmfR (6405 bp) was prepared prior to the
start of this project and provided by Dr. Christophe Corre. Most of the MmfR mutants
were produced by using Q5 site-directed mutagenesis kit (New England Biolabs)
according to the manufacturer’s guidelines with the primers used for each mutant shown
in Table 6.2. The elongation time for the PCR in site-directed mutagenesis was 3.5 mins.
The MmfR mutants Y85A and Y144A were prepared at the same time and provided by
Kathryn Styles from the Corre group (QuikChange Lightning, Agilent). All constructs
were sequence verified using the T7 primer pairs and transformed into chemically
competent E. coli BL21 star (DE3) or E. coli BL21 star (DE3)pLysS cells as described
in Section 6.3.7.
Table 6.2. Primers and conditions used for site-directed mutagenesis of MmfR.
MmfR mutant Primers Annealing temp.
Y85F For 5’-GAGGAGCACTTCGCGCGCTGG-3’
Rev 5’-CACCACGGCGATGGCCAG-3’ 72 °C
Y144F For 5’-CCCCTGCCCTTCGTGGACTGG-3’
Rev 5’-CAGCTCCGCGTCGATGAAG-3’ 69 °C
Q130E For 5’-TGCCCGGCTGGAGAGTGAGCG-3’
Rev 5’-CCGGCCTGCATCACGGGG-3’ 72 °C
L110G For 5’-CGAGGAGATGGGCCATCGCGCG-3’
Rev 5’-ACCGTCTCCAGCGGTGTG-3’ 67 °C
L110A For 5’-CGAGGAGATGGCCCATCGCGCG-3’
Rev 5’-ACCGTCTCCAGCGGTGTG-3’ 66 °C
L110V For 5’-CGAGGAGATGGTCCATCGCGC-3’
Rev 5’-ACCGTCTCCAGCGGTGTG-3’ 71 °C
Chapter 6 : Experimental
187
6.3.15 Expression and overproduction of recombinant protein
10 μL of glycerol stock of E. coli BL21star or E. coli BL21 star (DE3)pLysS cells
carrying the plasmid of interest was used to inoculate 10 mL LB medium containing
100 μg/mL of ampicillin and incubated for 16 hours (37 °C, 180 rpm). 5 mL of this
preculture was used to inoculate 1 L LB medium (0.5% v/v) containing 100 μg/mL of
ampicillin and grown (37 °C, 180 rpm) for 3-5 hours until optical density (OD) of cells
had reached 0.6-0.8 at 600 nm wavelength (Thermo BioMate3 spectrophotometer), when
isopropyl β-D-1-thiogalactopyranoside (ITPG, 5 mM stock solution) was added to a final
concentration of 0.5 mM. After addition of IPTG, the culture was grown for a further
16 hours (15 °C, 180 rpm).
6.3.16 Purification of recombinant protein
After incubation, cells were pelleted by centrifugation at 5000 rpm for 15 mins and
resuspended in washing buffer (15 mL per litre culture) containing 20 mM Tris-HCl
(pH 8), 100 mM NaCl, 20 mM imidazole and 10% v/v glycerol. Phenylmethanesuphonyl
fluoride (PMSF) was also added to give a final concentration of 1 mM. Resuspended
cells were stored at -80 °C until time of use. The resuspended cells were lysed using a
Constant Systems E1061 cell disrupter, and the cell debris pelleted by centrifugation at
15,000 rpm for 30 mins. A small quantity of the supernatant was diluted in 1 mL of
washing buffer and used as a cell-free extract containing the soluble proteins. Meanwhile,
a pipette tip was used to transfer a small quantity of the cell debris to 1 mL of washing
buffer, and the resuspended cell debris was used later as a sample containing the
insoluble proteins.
The supernatant was filtered (0.2 μm) and purified on an Äcta purifier (Amersham
Biosciences, GE Healthcare) at 4 °C or using a simple bench-top purification with a 1 mL
HisTrap HP nickel affinity column (GE Healthcare). In the bench-top purification, the
column was first equilibrated with washing buffer and loaded with the supernatant.
Unbound proteins were removed by washing with 5 mL of the above washing buffer and
His6-tagged protein was eluted using 3 mL of each elution buffer with successively
increasing amount of imidazole [20 mM Tris-HCl (pH 8), 100 mM NaCl, 10% v/v
glycerol and 50 mM, 100 mM, 200 mM or 300 mM imidazole]. The factions were
analysed by SDS-PAGE. The factions containing target protein were collected and
Chapter 6 : Experimental
188
concentrated using a Vivaspin ultrafiltration column (GE Healthcare or Sartorius) and
buffer-changed to the imidazole-free storage buffer containing 20 mM Tris-HCl (pH 8),
100 mM NaCl and 10% v/v glycerol. The protein was stored on ice, distributed into
approximately 50 μL aliquots, flash frozen in liquid nitrogen and stored at -80 °C.
6.3.17 Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE)
Samples were prepared by addition of 6 μL of 2X SDS-PAGE sample to 6 μL sample
and heated at 95 °C for 5 mins before loading onto the prepared gel. The 2X SDS-PAGE
sample buffer was prepared as follows: 0.5 M Tris-HCl, pH 6.8 (2.5 mL), glycerol
(2.0 mL), β-mercaptoethanol (0.4 mL), bromophenol blue (0.02 g) and SDS (0.4 g)
dissolved in distilled water to make a total volume of 10 mL, and then aliquoted and
frozen at -20 °C until needed.
15% SDS-PAGE gels were used for analysis of recombinant MmfR protein and its
mutants as well as SgnR protein, while 12% gel used for recombinant MmfL protein.
The resolving and stacking gels were prepared according to the recipe shown in Table
6.3. Tris buffer pH 8.8, Tris buffer pH 6.8 and tris-glycine SDS-PAGE running buffer
were prepared according to the procedure laid out by Sambrook and Russel (2001).204
Constituents were mixed and applied to the gel mould. After allowing the gel to set for
30 mins, samples were loaded and the gel run at 180 V until the dye front had reached
the bottom of the gel. All samples were run concurrently with PageRuler Plus prestained
protein ladder (Thermo Scientific). Protein staining of SDS-PAGE gels was performed
with Instant Blue (Expedeon).
Table 6.3. Recipe for 15% and 12% SDS-PAGE gels.
Constituents 15% Resolving gel
solution
12% Resolving gel
solution
5% Stacking gel
solution
Distilled water (mL) 1.1 1.6 0.68
29:1 (30%) Acrylamide : bisacrylamide
(mL) 2.5 2.0 0.17
Tris buffer (mL) 1.3 (pH 8.8, 1.5 M) 1.3 (pH 8.8, 1.5 M) 0.13 (pH 6.8, 1.0 M)
10% SDS solution (mL) 0.05 0.05 0.01
10% Ammonium persulfate solution (mL) 0.05 0.05 0.01
N,N,N,N-Tetramethylethylenediamine
(TEMED) (mL) 0.002 0.002 0.001
Total volume (mL) 5 5 1
Chapter 6 : Experimental
189
6.3.18 Electrophoretic mobility shift assay (EMSA)
Most of DNA sequences containing the operator for MmfR or SgnR used in EMSA
experiments were amplified by PCR reactions as described in Section 6.3.10. For hairpin
DNA, the ordered single-stranded oligonucleotides were diluted in sterile water to
200 nM and warmed at 95 °C for 2 mins and slowly cooled down to promote hairpin
formation (Table 6.4). For short annealed DNA fragments, paired single-stranded
oligonucleotides were diluted in sterile water to 400 nM each and equal volume of them
mixed together before heating at 95 °C for 5 mins and slowly cooling down (Table 6.4).
Table 6.4. Oligonucleotides for forming hairpin DNA and short annealed DNA
fragments.
Target Oligonucleotides Purpose
gbnA28 5’-AGTTGATACCTTCGTGAAGGTTTTTATT-3’
5’-AATAAAAACCTTCACGAAGGTATCAACT-3’
For preparing annealed 28-bp DNA containing the
operator for SgnR upstream of gbnA
gbnR-sgnH28 5’-AGAAAATACCTTCCAGACCGAATCTTTT-3’
5’-AAAAGATTCGGTCTGGAAGGTATTTTCT-3’
For preparing annealed 28-bp DNA containing the
operator for SgnR between gbnR and sgnH
Hairpin mmfRL 5’-GCAATATACCTGCGGGAAGGTATTATGCGAG
GCATAATACCTTCCCGCAGGTATATTGC-3’
For preparing hairpin DNA containing the operator
for MmfR between mmfR and mmfL
Hairpin mmyBY 5’-GCAAAAAACCTTCGGGAAGGTTTGACGCGAG
GCGTCAAACCTTCCCGAAGGTTTTTTGC-3’
For preparing hairpin DNA containing the operator
for MmfR between mmyY and mmyB
Hairpin mmyR 5’-GCAACATACCCTCGGGAAGGTATGTTGCGAG
GCAACATACCTTCCCGAGGGTATGTTGC-3’
For preparing hairpin DNA containing the operator
for MmfR upstream of mmyR
EMSAs were performed according to the manufacturer’s instructions in DIG gel shift kit,
2nd generation (Roche) with some modifications. Binding buffer and loading buffer were
prepared according to the recipe in the kit protocol. 5X binding buffer contains 100 mM
4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) (pH 7.6), 5 mM
ethylenediaminetetraacetic acid (EDTA), 50 mM (NH4)SO4, 5 mM dithiothreitol (DTT),
1% w/v Tween 20 and 150 mM KCl in distilled water. Loading buffer contains 34%
glycerol and 0.2% w/v bromophenol blue in 0.25X TBE buffer. 6% and 10% non-
denaturing polyacrylamide gels were freshly prepared as shown in Table 6.5 before use.
Most EMSAs were run on 6% non-denaturing polyacrylamide gels, while those with
hairpin DNA or short annealed DNA fragments were run on 10% gel due to their smaller
size. Constituents were mixed together and added to a gel mould with an appropriate
comb. Once set, gel was pre-run in 1X TBE buffer at 150 V at 4 °C for 1 hour.
A general reaction procedure is as follows. To a PCR tube on ice was added DNA, protein,
5X binding buffer and distilled water. Sample was incubated at room temperature for
Chapter 6 : Experimental
190
15 mins. Inducer dissolved in DMSO with appropriate concentrations was added to the
reaction tube, and the mixture incubated at room temperature for a further 10 mins. The
sample was then placed on ice and loading buffer was added to give a final volume of
25 μL before being loaded onto an equilibrated 6% non-denaturing polyacrylamide gel.
The gel was run at 80 V at 4 °C until the dye had migrated to the bottom of the gel. After
electrophoresis, the gel was stained in 100 mL of 1X TBE buffer with 5 μL GelRed at
room temperature for 30 mins with agitation before visualisation on a UV
transilluminator. A typical run is shown in Table 6.6. Total volume of each reaction did
not exceed 25 μL.
Table 6.5. Recipe for 6% and 10% non-denaturing polyacrylamide gels.
Constituents Non-denaturing polyacrylamide gel
6% 10%
Distilled water (mL) 5.97 4.64
29:1 (30%) Acrylamide: bisacrylamide (mL) 2.0 3.33
5X TBE buffer (mL) 2.0 2.0
10% Ammonium persulfate solution (mL) 0.07 0.07
TEMED (mL) 0.0035 0.0035
Total volume (mL) 10 10
Table 6.6. A typical run of an EMSA assay.
Reaction 1 2 3 4 5 6 7 8
DNA (μL) 4 4 4 4 4 4 4 4
Protein (μL) - 2 2 2 2 2 2 2
Binding buffer (μL) 4 4 4 4 4 4 4 4
Distilled water (μL) 12 10 6 6 6 6 6 6
Incubation at room temperature for 15 mins
Inducer (μL) with
increasing concentration - - 4 4 4 4 4 4
Incubation at room temperature for a further 10 mins
Loading buffer (μL) 5 5 5 5 5 5 5 5
Total volume (μL) 25 25 25 25 25 25 25 25
6.3.19 Enzymatic assay of MmfL
400 μM of S-(2-acetamidoethyl) 5-methyl-3-oxohexanethioate or S-(2-acetamidoethyl)
3-oxooctanethioate, 400 μM of dihydroxyacetone phophate and 3.0 nmol of recombinant
His6-MmfL protein in storage buffer with a final volume of 200 μL were incubated for
90 mins at room temperature. Then, 1 unit of shrimp alkaline phosphatase (New England
Chapter 6 : Experimental
191
Biolabs) was added and a further 45-min incubation at room temperature was carried out.
On completion of the reaction, 200 μL of methanol was added to the enzymatic mixture
and allowed to stand for 5 mins. After centrifugation at 14, 200 rpm for 10 mins, the
supernatant was filtered (0.4 μm) for LC-MS analyses. The amounts of protein used in
assays were tested from the standard reaction. For assay with denatured MmfL, protein
solution was boiled for 10 mins and cooled down before use.
References
192
References
1. Hopwood, D. A., Streptomyces in nature and medicine: the antibiotic makers. Oxford
University Press: 2007.
2. Poon, V. PhD thesis. University of Warwick, 2015.
3. McCormick, J. R.; Flärdh, K., FEMS Microbiol. Rev. 2012, 36 (1), 206-31.
4. Barka, E. A.; Vatsa, P.; Sanchez, L.; Gaveau-Vaillant, N.; Jacquard, C.; Klenk, H. P.;
Clément, C.; Ouhdouch, Y.; Van Wezeld, G. P., Microbiol. Mol. Biol. Rev. 2016, 80 (1), 1-43.
5. Al-Bassam, M. M.; Bibb, M. J.; Bush, M. J.; Chandra, G.; Buttner, M. J., PLoS Genet.
2014, 10 (8), e1004554.
6. Molle, V.; Buttner, M. J., Mol. Microbiol. 2000, 36 (6), 1265-78.
7. Waksman, S. A.; Woodruff, H. B., J. Bacteriol. 1942, 44 (3), 373-84.
8. Schatz, A.; Waksman, S. A., Exp. Biol. Med. 1944, 57 (2), 244-8.
9. Singh, B.; Mitchison, D. A., Br. Med. J. 1954, 1 (4854), 130-2.
10. Weitzman, D.; de Wend Cayley, F. E.; Wingfield, A. L., Respir. Med. 1950, 44 (4), 98-
104.
11. Finken, M.; Kirschner, P.; Meier, A.; Wrede, A.; Bottger, E. C., Mol. Microbiol. 1993,
9 (6), 1239-46.
12. de Lima Procópio, R. E.; da Silva, I. R.; Martins, M. K.; de Azevedo, J. L.; de Araújo, J.
M., Braz. J. Infect. Dis. 2012, 16 (5), 466-71.
13. Payne, D. J.; Gwynn, M. N.; Holmes, D. J.; Pompliano, D. L., Nat. Rev. Drug Discov.
2007, 6 (1), 29-40.
14. Alanis, A. J., Arch. Med. Res. 2005, 36 (6), 697-705.
15. Whiley, D. M.; Lahra, M. M.; Unemo, M., Future Microbiol. 2015, 10 (3), 313-6.
16. Watve, M. G.; Tickoo, R.; Jog, M. M.; Bhole, B. D., Arch. Microbiol. 2001, 176 (5),
386-90.
17. Katz, L.; Baltz, R. H., J. Ind. Microbiol. Biotechnol. 2016, 43 (2-3), 155-76.
18. Challis, G. L., J. Ind. Microbiol. Biotechnol. 2014, 41 (2), 219-32.
19. Bentley, S. D.; Chater, K. F.; Cerdeno-Tarraga, A. M.; Challis, G. L.; Thomson, N. R.;
James, K. D.; Harris, D. E.; Quail, M. A.; Kieser, H.; Harper, D.; Bateman, A.; Brown, S.;
Chandra, G.; Chen, C. W.; Collins, M.; Cronin, A.; Fraser, A.; Goble, A.; Hidalgo, J.; Hornsby,
T.; Howarth, S.; Huang, C. H.; Kieser, T.; Larke, L.; Murphy, L.; Oliver, K.; O'Neil, S.;
Rabbinowitsch, E.; Rajandream, M. A.; Rutherford, K.; Rutter, S.; Seeger, K.; Saunders, D.;
Sharp, S.; Squares, R.; Squares, S.; Taylor, K.; Warren, T.; Wietzorrek, A.; Woodward, J.; Barrell,
B. G.; Parkhill, J.; Hopwood, D. A., Nature 2002, 417 (6885), 141-7.
References
193
20. Bentley, S. D.; Brown, S.; Murphy, L. D.; Harris, D. E.; Quail, M. A.; Parkhill, J.; Barrell,
B. G.; McCormick, J. R.; Santamaria, R. I.; Losick, R.; Yamasaki, M.; Kinashi, H.; Chen, C. W.;
Chandra, G.; Jakimowicz, D.; Kieser, H. M.; Kieser, T.; Chater, K. F., Mol. Microbiol. 2004, 51
(6), 1615-28.
21. Haug, I.; Weissenborn, A.; Brolle, D.; Bentley, S.; Kieser, T.; Altenbuchner, J.,
Microbiology 2003, 149 (2), 505-13.
22. Davis, N. K.; Chater, K. F., Mol. Microbiol. 1990, 4 (10), 1679-91.
23. Wright, L. F.; Hopwood, D. A., J. Gen. Microbiol. 1976, 95 (1), 96-106.
24. Kirby, R.; Hopwood, D. A., J. Gen. Microbiol. 1977, 98 (1), 239-52.
25. Hornemann, U.; Hopwood, D. A., Biosynthesis of Methylenomycin A: A Plasmid-
Determined Antibiotic. In Biosynthesis, Corcoran, J. W., Ed. Springer Berlin Heidelberg: Berlin,
Heidelberg, 1981; pp 123-31.
26. Wright, L. F.; Hopwood, D. A., J. Gen. Microbiol. 1976, 96 (2), 289-97.
27. Haynes, S. W.; Sydor, P. K.; Corre, C.; Song, L.; Challis, G. L., J. Am. Chem. Soc. 2011,
133 (6), 1793-8.
28. Hopwood, D. A.; Wright, H. M., J. Gen. Microbiol. 1983, 129 (12), 3575-9.
29. Gottelt, M.; Kol, S.; Gomez-Escribano, J. P.; Bibb, M.; Takano, E., Microbiology 2010,
156 (8), 2343-53.
30. Gomez-Escribano, J. P.; Song, L.; Fox, D. J.; Yeo, V.; Bibb, M. J.; Challis, G. L., Chem.
Sci. 2012, 3 (9), 2716-20.
31. Gurtler, H.; Pedersen, R.; Anthoni, U.; Christophersen, C.; Nielsen, P. H.; Wellington,
E. M.; Pedersen, C.; Bock, K., J. Antibiot. 1994, 47 (4), 434-9.
32. Zhao, B.; Lin, X.; Lei, L.; Lamb, D. C.; Kelly, S. L.; Waterman, M. R.; Cane, D. E., J.
Biol. Chem. 2008, 283 (13), 8183-9.
33. Challis, G. L.; Ravel, J., FEMS Microbiol. Lett. 2000, 187 (2), 111-4.
34. Lautru, S.; Deeth, R. J.; Bailey, L. M.; Challis, G. L., Nat. Chem. Biol. 2005, 1 (5), 265-
9.
35. Barona-Gomez, F.; Wong, U.; Giannakopulos, A. E.; Derrick, P. J.; Challis, G. L., J. Am.
Chem. Soc. 2004, 126 (50), 16282-3.
36. Ohnishi, Y.; Ishikawa, J.; Hara, H.; Suzuki, H.; Ikenoya, M.; Ikeda, H.; Yamashita, A.;
Hattori, M.; Horinouchi, S., J. Bacteriol. 2008, 190 (11), 4050-60.
37. Ikeda, H.; Ishikawa, J.; Hanamoto, A.; Shinose, M.; Kikuchi, H.; Shiba, T.; Sakaki, Y.;
Hattori, M.; Omura, S., Nat. Biotechnol. 2003, 21 (5), 526-31.
38. Pullan, S. T.; Chandra, G.; Bibb, M. J.; Merrick, M., BMC genomics 2011, 12, 175.
39. Zerikly, M.; Challis, G. L., ChemBioChem 2009, 10 (4), 625-33.
40. Medema, M. H.; Breitling, R.; Bovenberg, R.; Takano, E., Nat. Rev. Microbiol. 2011, 9
(2), 131-7.
References
194
41. Rutledge, P. J.; Challis, G. L., Nat. Rev. Microbiol. 2015, 13 (8), 509-23.
42. Gomez-Escribano, J. P.; Bibb, M. J., Microb. Biotechnol. 2011, 4 (2), 207-15.
43. Camilli, A.; Bassler, B. L., Science 2006, 311 (5764), 1113-6.
44. Takano, E., Curr. Opin. Microbiol. 2006, 9 (3), 287-94.
45. Corre, C.; Song, L.; O'Rourke, S.; Chater, K. F.; Challis, G. L., Proc. Natl. Acad. Sci. U.
S. A. 2008, 105 (45), 17510-5.
46. Khokhlov, A. S.; Tovarova, I. I.; Borisova, L. N.; Pliner, S. A.; Shevchenko, L. N.;
Kornitskaia, E. I.; Ivkina, N. S.; Rapoport, I. A., Doklady Akademii nauk SSSR 1967, 177 (1),
232-5.
47. Hara, O.; Beppu, T., J. Antibiot. 1982, 35 (3), 349-58.
48. Ohnishi, Y.; Furusho, Y.; Higashi, T.; Chun, H. K.; Furihata, K.; Sakuda, S.; Horinouchi,
S., J. Antibiot. 2004, 57 (3), 218-23.
49. Horinouchi, S.; Beppu, T., Mol. Microbiol. 1994, 12 (6), 859-64.
50. Hsiao, N.-H.; Nakayama, S.; Merlo, M. E.; de Vries, M.; Bunet, R.; Kitani, S.; Nihira,
T.; Takano, E., Chem. Biol. 2009, 16 (9), 951-60.
51. Kitani, S.; Kinoshita, H.; Nihira, T.; Yamada, Y., J. Bacteriol. 1999, 181 (16), 5081-4.
52. Gräfe, U.; Schade, W.; Eritt, I.; Fleck, W.; Radics, L., J. Antibiot. 1982, 35 (12), 1722-
3.
53. Sakuda, S.; Yamada, Y., Tetrahedron Lett. 1991, 32 (15), 1817-20.
54. Gräfe, U.; Reinhardt, G.; Schade, W.; Eritt, I.; Fleck, W. F.; Radics, L., Biotechnol. Lett.
1983, 5 (9), 591-6.
55. Yamada, Y.; Sugamura, K.; Kondo, K.; Yanagimoto, M.; Okada, H., J. Antibiot. 1987,
40 (4), 496-504.
56. Kondo, K.; Higuchi, Y.; Sakuda, S.; Nihira, T.; Yamada, Y., J. Antibiot. 1989, 42 (12),
1873-6.
57. Takano, E.; Nihira, T.; Hara, Y.; Jones, J. J.; Gershater, C. J.; Yamada, Y.; Bibb, M., J.
Biol. Chem. 2000, 275 (15), 11010-6.
58. O'Rourke, S.; Wietzorrek, A.; Fowler, K.; Corre, C.; Challis, G. L.; Chater, K. F., Mol.
Microbiol. 2009, 71 (3), 763-78.
59. Sidda, J. D.; Song, L.; Poon, V.; Al-Bassam, M.; Lazos, O.; Buttner, M. J.; Challis, G.
L.; Corre, C., Chem. Sci. 2014, 5 (1), 86-9.
60. Sidda, J. D. PhD thesis. University of Warwick, 2015.
61. Kitani, S.; Miyamoto, K. T.; Takamatsu, S.; Herawati, E.; Iguchi, H.; Nishitomi, K.;
Uchida, M.; Nagamitsu, T.; Omura, S.; Ikeda, H.; Nihira, T., Proc. Natl. Acad. Sci. U. S. A. 2011,
108 (39), 16410-5.
62. Arakawa, K.; Tsuda, N.; Taniguchi, A.; Kinashi, H., ChemBioChem 2012, 13 (10), 1447-
57.
References
195
63. Horinouchi, S.; Kumada, Y.; Beppu, T., J. Bacteriol. 1984, 158 (2), 481-7.
64. Horinouchi, S.; Suzuki, H.; Nishiyama, M.; Beppu, T., J. Bacteriol. 1989, 171 (2), 1206-
10.
65. Horinouchi, S.; Nishiyama, M.; Suzuki, H.; Kumada, Y.; Beppu, T., J. Antibiot. 1985,
38 (5), 636-41.
66. Sakuda, S.; Higashi, A.; Tanaka, S.; Nihira, T.; Yamada, Y., J. Am. Chem. Soc. 1992,
114 (2), 663-8.
67. Kato, J. Y.; Funa, N.; Watanabe, H.; Ohnishi, Y.; Horinouchi, S., Proc. Natl. Acad. Sci.
U. S. A. 2007, 104 (7), 2378-83.
68. Han, L.; Lobo, S.; Reynolds, K. A., J. Bacteriol. 1998, 180 (17), 4481-6.
69. Li, Y.; Florova, G.; Reynolds, K. A., J. Bacteriol. 2005, 187 (11), 3795-9.
70. Hsiao, N.-H.; Söding, J.; Linke, D.; Lange, C.; Hertweck, C.; Wohlleben, W.; Takano,
E., Microbiology 2007, 153 (5), 1394-404.
71. Takano, E.; Chakraburtty, R.; Nihira, T.; Yamada, Y.; Bibb, M. J., Mol. Microbiol. 2001,
41 (5), 1015-28.
72. Kawachi, R.; Akashi, T.; Kamitani, Y.; Sy, A.; Wangchaisoonthorn, U.; Nihira, T.;
Yamada, Y., Mol. Microbiol. 2000, 36 (2), 302-13.
73. Waki, M.; Nihira, T.; Yamada, Y., J. Bacteriol. 1997, 179 (16), 5131-7.
74. Shikura, N.; Yamamura, J.; Nihira, T., J. Bacteriol. 2002, 184 (18), 5151-7.
75. Corre, C.; Haynes, S. W.; Malet, N.; Song, L.; Challis, G. L., Chem. Commun. 2010, 46
(23), 4079-81.
76. Davis, J. B.; Bailey, J. D.; Sello, J. K., Org. Lett. 2009, 11 (14), 2984-7.
77. Malet, N. PhD thesis. University of Warwick, 2012.
78. Cuthbertson, L.; Nodwell, J. R., Microbiol. Mol. Biol. Rev. 2013, 77 (3), 440-75.
79. Yang, H. L.; Zubay, G.; Levy, S. B., Proc. Natl. Acad. Sci. U. S. A. 1976, 73 (5), 1509-
12.
80. Orth, P.; Schnappinger, D.; Hillen, W.; Saenger, W.; Hinrichs, W., Nat. Struct. Mol. Biol.
2000, 7 (3), 215-9.
81. Retzlaff, L.; Distler, J., Mol. Microbiol. 1995, 18 (1), 151-62.
82. Vujaklija, D.; Horinouchi, S.; Beppu, T., J. Bacteriol. 1993, 175 (9), 2652-61.
83. Hsiao, N.-H.; Gottelt, M.; Takano, E., Methods Enzymol. 2009, 458, 143-57.
84. Bibb, M. J., Curr. Opin. Microbiol. 2005, 8 (2), 208-15.
85. Xu, G.; Wang, J.; Wang, L.; Tian, X.; Yang, H.; Fan, K.; Yang, K.; Tan, H., J. Biol.
Chem. 2010, 285 (35), 27440-8.
86. Haneishi, T.; Terahara, A.; Arai, M.; Hata, T.; Tamura, C., J. Antibiot. 1974, 27 (6), 393-
9.
References
196
87. Folcher, M.; Gaillard, H.; Nguyen, L. T.; Nguyen, K. T.; Lacroix, P.; Bamas-Jacques,
N.; Rinkel, M.; Thompson, C. J., J. Biol. Chem. 2001, 276 (47), 44297-306.
88. Leskiw, B. K.; Mah, R.; Lawlor, E. J.; Chater, K. F., J. Bacteriol. 1993, 175 (7), 1995-
2005.
89. Sah, S.; Singh, R., Agriculture 2015, 61 (3), 97-114.
90. Saha, M.; Sarkar, S.; Sarkar, B.; Sharma, B. K.; Bhattacharjee, S.; Tribedi, P., Environ.
Sci. Pollut. Res. Int. 2016, 23 (5), 3984-99.
91. Sheldon, J. R.; Heinrichs, D. E., FEMS Microbiol. Rev. 2015, 39 (4), 592-630.
92. Ratledge, C.; Dover, L. G., Annu. Rev. Microbiol. 2000, 54, 881-941.
93. Crowley, D. E.; Wang, Y. C.; Reid, C. P. P.; Szaniszlo, P. J., Plant Soil 1991, 130 (1),
179-98.
94. Cornish, A. S.; Page, W. J., BioMetals 1995, 8 (4), 332-8.
95. Corbin, J. L.; Bulen, W. A., Biochemistry 1969, 8 (3), 757-62.
96. Page, W. J.; Tigerstrom, M. V., Microbiology 1988, 134 (2), 453-60.
97. Wittmann, S.; Heinisch, L.; Scherlitz-Hofmann, I.; Stoiber, T.; Ankel-Fuchs, D.;
Möllmann, U., BioMetals 2004, 17 (1), 53-64.
98. Ward, T. R.; Lutz, A.; Parel, S. P.; Ensling, J.; Gutlich, P.; Buglyo, P.; Orvig, C., Inorg.
Chem. 1999, 38 (22), 5007-17.
99. Raymond, K. N.; Dertz, E. A.; Kim, S. S., Proc. Natl. Acad. Sci. U. S. A. 2003, 100 (7),
3584-8.
100. Münzinger, M.; Taraz, K.; Budzikiewicz, H.; Drechsel, H.; Heymann, P.; Winkelmann,
G.; Meyer, J.-M., BioMetals 1999, 12 (2), 189-93.
101. De Voss, J. J.; Rutter, K.; Schroeder, B. G.; Su, H.; Zhu, Y.; Barry, C. E., Proc. Natl.
Acad. Sci. U. S. A. 2000, 97 (3), 1252-7.
102. Lytton, S. D.; Mester, B.; Dayan, I.; Glickstein, H.; Libman, J.; Shanzer, A.; Cabantchik,
Z. I., Blood 1993, 81 (1), 214-21.
103. Mislin, G. L. A.; Schalk, I. J., Metallomics 2014, 6 (3), 408-20.
104. Barry, S. M.; Challis, G. L., Curr. Opin. Chem. Biol. 2009, 13 (2), 205-15.
105. Crosa, J. H.; Walsh, C. T., Microbiol. Mol. Biol. Rev. 2002, 66 (2), 223-49.
106. Guerinot, M. L., Annu. Rev. Microbiol. 1994, 48, 743-72.
107. Miethke, M.; Marahiel, M. A., Microbiol. Mol. Biol. Rev. 2007, 71 (3), 413-51.
108. Actis, L. A.; Fish, W.; Crosa, J. H.; Kellerman, K.; Ellenberger, S. R.; Hauser, F. M.;
Sanders-Loehr, J., J. Bacteriol. 1986, 167 (1), 57-65.
109. Jalal, M. A. F.; Hossain, M. B.; Van der Helm, D.; Sanders-Loehr, J.; Actis, L. A.; Crosa,
J. H., J. Am. Chem. Soc. 1989, 111 (1), 292-6.
110. Drechsel, H.; Stephan, H.; Lotz, R.; Haag, H.; Zähner, H.; Hantke, K.; Jung, G., Liebigs
Ann. Recl. 1995, 1995 (10), 1727-33.
References
197
111. Cox, C. D.; Rinehart, K. L.; Moore, M. L.; Cook, J. C., Proc. Natl. Acad. Sci. U. S. A.
1981, 78 (7), 4256-60.
112. Zunnundzhanov, A.; Bessonova, I. A.; Abdullaev, N. D.; Ogai, D. K., Chem. Nat. Compd.
1987, 23 (4), 461-5.
113. Lee, J. Y.; Moon, S. S.; Hwang, B. K., Appl. Environ. Microbiol. 2003, 69 (4), 2023-31.
114. Youard, Z. A.; Mislin, G. L.; Majcherczyk, P. A.; Schalk, I. J.; Reimmann, C., J. Biol.
Chem. 2007, 282 (49), 35546-53.
115. Lin, Z.; Antemano, R. R.; Hughen, R. W.; Tianero, M. D. B.; Peraud, O.; Haygood, M.
G.; Concepcion, G. P.; Olivera, B. M.; Light, A.; Schmidt, E. W., J. Nat. Prod. 2010, 73 (11),
1922-6.
116. Shindo, K.; Takenaka, A.; Noguchi, T.; Hayakawa, Y.; Seto, H., J. Antibiot. 1989, 42
(10), 1526-9.
117. Sasaki, O.; Igarashi, Y.; Saito, N.; Furumai, T., J. Antibiot. 2002, 55 (3), 249-55.
118. Igarashi, Y.; Asano, D.; Sawamura, M.; In, Y.; Ishida, T.; Imoto, M., Org. Lett. 2016, 18
(7), 1658-61.
119. Thomas, M. S., BioMetals 2007, 20 (3), 431-52.
120. Castignetti, D., Curr. Microbiol. 1997, 34 (4), 250-7.
121. Seipke, R. F.; Song, L.; Bicz, J.; Laskaris, P.; Yaxley, A. M.; Challis, G. L.; Loria, R.,
Microbiology 2011, 157 (9), 2681-93.
122. Klumpp, C.; Burger, A.; Mislin, G. L.; Abdallah, M. A., Bioorg. Med. Chem. Lett. 2005,
15 (6), 1721-4.
123. Miller, M. C.; Parkin, S.; Fetherston, J. D.; Perry, R. D.; DeMoll, E., J. Inorg. Biochem.
2006, 100 (9), 1495-500.
124. Budzikiewicz, H., Fortschr. Chem. Org. Naturst. 2004, 87, 81-237.
125. Meyer, J. M., Arch. Microbiol. 2000, 174 (3), 135-42.
126. Adler, C.; Corbalán, N. S.; Seyedsayamdost, M. R.; Pomares, M. F.; de Cristóbal, R. E.;
Clardy, J.; Kolter, R.; Vincent, P. A., PLoS ONE 2012, 7 (10), e46754.
127. Noel, S.; Gasser, V.; Pesset, B.; Hoegy, F.; Rognan, D.; Schalk, I. J.; Mislin, G. L. A.,
Org. Biomol. Chem. 2011, 9 (24), 8288-300.
128. Quadri, L. E. N.; Keating, T. A.; Patel, H. M.; Walsh, C. T., Biochemistry 1999, 38 (45),
14941-54.
129. Patel, H. M.; Walsh, C. T., Biochemistry 2001, 40 (30), 9023-31.
130. Serino, L.; Reimmann, C.; Baur, H.; Beyeler, M.; Visca, P.; Haas, D., Mol. Gen. Genet.
1995, 249 (2), 217-28.
131. Serino, L.; Reimmann, C.; Visca, P.; Beyeler, M.; Chiesa, V. D.; Haas, D., J. Bacteriol.
1997, 179 (1), 248-57.
References
198
132. Reimmann, C.; Patel, H. M.; Serino, L.; Barone, M.; Walsh, C. T.; Haas, D., J. Bacteriol.
2001, 183 (3), 813-20.
133. Reimmann, C.; Patel, H. M.; Walsh, C. T.; Haas, D., J. Bacteriol. 2004, 186 (19), 6367-
73.
134. Kotowska, M.; Pawlik, K., Appl. Microbiol. Biotechnol. 2014, 98 (18), 7735-46.
135. Youard, Z. A.; Wenner, N.; Reimmann, C., BioMetals 2011, 24 (3), 513-22.
136. Inahashi, Y.; Zhou, S.; Bibb, M. J.; Song, L.; Al-Bassam, M. M.; Bibb, M. J.; Challis, G.
L., Chem. Sci. 2016, submitted for publication.
137. Fontecave, M.; Atta, M.; Mulliez, E., Trends Biochem. Sci. 2004, 29 (5), 243-9.
138. Dong, C.; Huang, F.; Deng, H.; Schaffrath, C.; Spencer, J. B.; O'Hagan, D.; Naismith, J.
H., Nature 2004, 427 (6974), 561-5.
139. Deng, H.; Cobb, S. L.; McEwan, A. R.; McGlinchey, R. P.; Naismith, J. H.; O'Hagan,
D.; Robinson, D. A.; Spencer, J. B., Angew. Chem. Int. Ed. (English) 2006, 45 (5), 759-62.
140. Lin, H., Bioorg. Chem. 2011, 39 (5-6), 161-70.
141. Bauerle, M. R.; Schwalm, E. L.; Booker, S. J., J. Biol. Chem. 2015, 290 (7), 3995-4002.
142. Broderick, J. B.; Duffus, B. R.; Duschene, K. S.; Shepard, E. M., Chem. Rev. 2014, 114
(8), 4229-317.
143. Zhang, Q.; van der Donk, W. A.; Liu, W., Acc. Chem. Res. 2012, 45 (4), 555-64.
144. Hu, Y.; Ribbe, M. W., Curr. Opin. Chem. Biol. 2016, 31, 188-94.
145. Benítez-Páez, A.; Villarroya, M.; Armengod, M.-E., RNA 2012, 18 (10), 1783-95.
146. Giessing, A. M. B.; Jensen, S. S.; Rasmussen, A.; Hansen, L. H.; Gondela, A.; Long, K.;
Vester, B.; Kirpekar, F., RNA 2009, 15 (2), 327-36.
147. Allen, K. D.; Xu, H.; White, R. H., J. Bacteriol. 2014, 196 (18), 3315-23.
148. Zhou, S.; Alkhalaf, L.; de los Santos, E. L. C.; Challis, G. L., Curr. Opin. Chem. Biol.
2016, 35, 73-79.
149. Huang, C.; Huang, F.; Moison, E.; Guo, J.; Jian, X.; Duan, X.; Deng, Z.; Leadlay, P. F.;
Sun, Y., Chem. Biol. 2015, 22 (2), 251-61.
150. Kuzuyama, T.; Seki, T.; Dairi, T.; Hidaka, T.; Seto, H., J. Antibiot. 1995, 48 (10), 1191-
3.
151. Marous, D. R.; Lloyd, E. P.; Buller, A. R.; Moshos, K. A.; Grove, T. L.; Blaszczyk, A.
J.; Booker, S. J.; Townsend, C. A., Proc. Natl. Acad. Sci. U. S. A. 2015, 112 (33), 10354-8.
152. Rachid, S.; Scharfe, M.; Blöcker, H.; Weissman, K. J.; Müller, R., Chem. Biol. 2009, 16
(1), 70-81.
153. Benjdia, A.; Pierre, S.; Gherasim, C.; Guillot, A.; Carmona, M.; Amara, P.; Banerjee, R.;
Berteau, O., Nat. Commun. 2015, 6, 8377.
154. Blaszczyk, A. J.; Silakov, A.; Zhang, B.; Maiocco, S. J.; Lanz, N. D.; Kelly, W. L.; Elliott,
S. J.; Krebs, C.; Booker, S. J., J. Am. Chem. Soc. 2016, 138 (10), 3416-26.
References
199
155. Huo, L.; Rachid, S.; Stadler, M.; Wenzel, S. C.; Muller, R., Chem. Biol. 2012, 19 (10),
1278-87.
156. Gomez-Escribano, J. P.; Song, L.; Bibb, M. J.; Challis, G. L., Chem. Sci. 2012, 3 (12),
3522-5.
157. Crone, W. J. K.; Leeper, F. J.; Truman, A. W., Chem. Sci. 2012, 3 (12), 3516-21.
158. Freeman, M. F.; Gurgui, C.; Helf, M. J.; Morinaka, B. I.; Uria, A. R.; Oldham, N. J.;
Sahl, H.-G.; Matsunaga, S.; Piel, J., Science 2012, 338 (6105), 387-90.
159. Werner, W. J.; Allen, K. D.; Hu, K.; Helms, G. L.; Chen, B. S.; Wang, S. C.,
Biochemistry 2011, 50 (42), 8986-8.
160. Ahlert, J.; Shepard, E.; Lomovskaya, N.; Zazopoulos, E.; Staffa, A.; Bachmann, B. O.;
Huang, K.; Fonstein, L.; Czisny, A.; Whitwam, R. E.; Farnet, C. M.; Thorson, J. S., Science 2002,
297 (5584), 1173-6.
161. Lohman, J. R.; Huang, S. X.; Horsman, G. P.; Dilfer, P. E.; Huang, T.; Chen, Y.; Wendt-
Pienkowski, E.; Shen, B., Mol. Biosyst. 2013, 9 (3), 478-91.
162. Westrich, L.; Heide, L.; Li, S. M., ChemBioChem 2003, 4 (8), 768-73.
163. Welander, P. V.; Coleman, M. L.; Sessions, A. L.; Summons, R. E.; Newman, D. K.,
Proc. Natl. Acad. Sci. U. S. A. 2010, 107 (19), 8537-42.
164. Noike, M.; Matsui, T.; Ooya, K.; Sasaki, I.; Ohtaki, S.; Hamano, Y.; Maruyama, C.;
Ishikawa, J.; Satoh, Y.; Ito, H.; Morita, H.; Dairi, T., Nat. Chem. Biol. 2015, 11 (1), 71-6.
165. Ostash, B.; Doud, E. H.; Lin, C.; Ostash, I.; Perlstein, D. L.; Fuse, S.; Wolpert, M.; Kahne,
D.; Walker, S., Biochemistry 2009, 48 (37), 8830-41.
166. Gilbert, I. H.; Ginty, M.; O'Neill, J. A.; Simpson, T. J.; Staunton, J.; Willis, C. L., Bioorg.
Med. Chem. Lett. 1995, 5 (15), 1587-90.
167. Georgiades, S. N.; Clardy, J., Bioorg. Med. Chem. Lett. 2008, 18 (10), 3117-21.
168. Morin, J. B.; Adams, K. L.; Sello, J. K., Org. Biomol. Chem. 2012, 10 (8), 1517-20.
169. Sakuda, S.; Tanaka, S.; Mizuno, K.; Sukcharoen, O.; Nihira, T.; Yamada, Y., J. Chem.
Soc., Perkin Trans. 1 1993, (19), 2309-15.
170. Hellman, L. M.; Fried, M. G., Nat. Protoc. 2007, 2 (8), 1849-61.
171. Laniel, M. A.; Beliveau, A.; Guerin, S. L., Methods Mol. Biol. 2001, 148, 13-30.
172. Stanton, B. C.; Nielsen, A. A. K.; Tamsir, A.; Clancy, K.; Peterson, T.; Voigt, C. A., Nat.
Chem. Biol. 2014, 10 (2), 99-105.
173. Mori, K., Tetrahedron Lett. 1981, 22 (35), 3431-2.
174. Posner, G. H.; Weitzberg, M.; Jew, S.-S., Synth. Commun. 1987, 17 (6), 611-20.
175. Kinoshita, T.; Hirano, M., J. Heterocycl. Chem. 1992, 29 (4), 1025-6.
176. Parsons, P. J.; Lacrouts, P.; Buss, A. D., J. Chem. Soc., Chem. Commun. 1995, (4), 437-
8.
References
200
177. Crawforth, J. M.; Fawcett, J.; Rawlings, B. J., J. Chem. Soc., Perkin Trans. 1 1998, (10),
1721-6.
178. Chavan, S. P.; Pasupathy, K.; Shivasankar, K., Synth. Commun. 2004, 34 (3), 397-404.
179. Fernández-Martínez, L. T.; Borsetto, C.; Gomez-Escribano, J. P.; Bibb, M. J.; Al-
Bassam, M. M.; Chandra, G.; Bibb, M. J., Antimicrob. Agents Chemother. 2014, 58 (12), 7441-
50.
180. Ankenbauer, R. G.; Toyokuni, T.; Staley, A.; Rinehart Jr, K. L.; Cox, C. D., J. Bacteriol.
1988, 170 (11), 5344-51.
181. Rinehart, K. L.; Staley, A. L.; Wilson, S. R.; Ankenbauer, R. G.; Cox, C. D., J. Org.
Chem. 1995, 60 (9), 2786-91.
182. Zamri, A.; Abdallah, M. A., Tetrahedron 2000, 56 (2), 249-56.
183. Dunetz, J. R.; Magano, J.; Weisenburger, G. A., Org. Process Res. Dev. 2016, 20 (2),
140-77.
184. Al-Warhi, T. I.; AL-Hazimi, H. M. A.; El-Faham, A.; Albericio, F., Molecules 2010, 15
(12), 9403-17.
185. Allen, K. D.; Wang, S. C., Arch. Biochem. Biophys. 2014, 543, 67-73.
186. Kuzuyama, T.; Hidaka, T.; Kamigiri, K.; Imai, S.; Seto, H., J. Antibiot. 1992, 45 (11),
1812-4.
187. Woodyer, R. D.; Li, G.; Zhao, H.; van der Donk, W. A., Chem. Commun. 2007, (4),
359-61.
188. Hammerschmidt, F., Liebigs Ann. Chem. 1992, 1992 (6), 553-7.
189. Hammerschmidt, F.; Kaehlig, H., J. Org. Chem. 1991, 56 (7), 2364-70.
190. Kellenberger, J. L. PhD thesis. ETH Zurich, 1997.
191. Aberhart, D. J.; Lin, L. J.; Chu, J. Y.-R., J. Chem. Soc., Perkin Trans. 1 1975, (23),
2517-23.
192. Young, D. W.; Sen, P. K.; Morecombe, D. J., Eur. J. Biochem. 1977, 75 (1), 133-47.
193. Borcsok, E.; Abeles, R. H., Arch. Biochem. Biophys. 1982, 213 (2), 695-707.
194. Axelsson, B. S.; O'Toole, K. J.; Spencer, P. A.; Young, D. W., J. Chem. Soc., Perkin
Trans. 1 1994, (7), 807-15.
195. Ruszczycky, M. W.; Anderson, V. E., Bioorg. Chem. 2004, 32 (1), 51-61.
196. Ramer, S. E.; Moore, R. N.; Vederas, J. C., Can. J. Chem. 1986, 64 (4), 706-13.
197. Slieker, L.; Benkovic, S. J., J. Labelled Compd. Radiopharm. 1982, 19 (5), 647-57.
198. Baldwin, J. E.; Adlington, R. M.; Robinson, N. G.; Ting, H.-H., J. Chem. Soc., Chem.
Commun. 1986, (5), 409-11.
199. Strauss, E.; Begley, T. P., Bioorg. Med. Chem. Lett. 2003, 13 (3), 339-42.
200. Maeda, Y.; Tago, K.; Eguchi, T.; Kakinuma, K., Biosci., Biotechnol., Biochem. 1996, 60
(8), 1248-54.
References
201
201. Oba, M.; Iwasaki, A.; Hitokawa, H.; Ikegame, T.; Banba, H.; Ura, K.; Takamura, T.;
Nishiyama, K., Tetrahedron: Asymmetry 2006, 17 (12), 1890-4.
202. Jurczak, J.; Golebiowski, A., Chem. Rev. 1989, 89 (1), 149-64.
203. Eliot, A. C.; Kirsch, J. F., Annu. Rev. Biochem 2004, 73, 383-415.
204. Sambrook, J.; Russell, D. W., Molecular cloning: a laboratory manual 3rd edition. Cold
Spring Harbor Laboratory Press: 2001.